210
IEEE JOURNAL OF QUANTUM ELECTRONICS, VOL. 48, NO. 2, FEBRUARY 2012
Photonic Crystal Silicon Optical Modulators:
Carrier-Injection and Depletion at 10 Gb/s
Hong C. Nguyen, Yuya Sakai, Mizuki Shinkawa, Norihiro Ishikura, and Toshihiko Baba, Member, IEEE
(Invited Paper)
Abstract— We demonstrate 10 Gb/s modulation in a 200 µm
photonic crystal silicon optical modulator, in both carrierinjection and depletion modes. In particular, this is the first
demonstration of 10 Gb/s modulation in depletion mode and
without pre-emphasis, in a Mach–Zehnder type modulator of
this length, although moderate pre-emphasis can improve the
signal quality. This is made possible by utilizing the slow-light
of the photonic crystal waveguide, where the group index ng
is ∼30 and gives ∼7 times enhancement in the modulation
efficiency compared to rib-waveguide devices. We observe 10 Gb/s
modulation at drive voltages as low as 1.6 V and 3.6 V peak-topeak, in injection- and depletion-modes, respectively.
Index Terms— Optical modulators,
waveguides, silicon photonics.
photonic
crystal
I. I NTRODUCTION
A
S A KEY component of silicon-photonics-based optical
interconnects, there is a strong demand for silicon optical
modulators that satisfy multiple criteria including small footprint as well as high-speed (>10 Gb/s), low-voltage/power,
low-loss and high extinction ratio, just to name a few [1].
Each of the different types of silicon optical modulators, based
on Mach-Zehnder interferometers (MZIs), μ-ring resonators
and electro-absorption for example, have their strengths and
shortfalls and none of them are yet to satisfy all criteria.
Although MZI-based modulators are considered versatile as
they are capable of both amplitude and phase modulation, as
well as having a large working spectrum, their main drawback
have been their long device lengths. Most MZI modulators,
particularly those that operate by carrier-depletion, have phaseshifter lengths of several millimeters and result in a large
device footprint [2-4]. Further reduction in their device lengths
to sub-millimeter scale is required for large-scale integration
with other optical components. One way to achieve this
is through modified p-n structures to increase the overlap
between carrier motion and the optical mode, thereby increasing the modulation efficiency [2, 5, 6]. However, it would be
difficult to achieve multi-fold increase in efficiency through
Manuscript received August 1, 2011; revised October 26, 2011; accepted
October 27, 2011. Date of current version January 24, 2012. This work was
supported in part by the Funding Program for World-Leading Innovative
Research and Development on Science and Technology (FIRST Program)
of Japan Society for the Promotion of Science.
The authors are with the Department of Electrical and Computer Engineering, Yokohama National University, Yokohama 240-8501,
Japan (e-mail: hong@ynu.ac.jp; d11gd145@ynu.ac.jp; d10gd144@ynu.ac.jp;
d09gd105@ynu.ac.jp; baba@ynu.ac.jp).
Color versions of one or more of the figures in this paper are available
online at http://ieeexplore.ieee.org.
Digital Object Identifier 10.1109/JQE.2011.2174338
this method. Another method, with which multi-fold efficiency
increase is possible, is to incorporate slow-light structures.
Photonic crystal waveguides (PCW) are one such structure
capable of generating slow-light. PCWs can easily exhibit a
group index (n g ) of 50 which is over 10 times larger than in
rib-waveguides often used in silicon MZI modulators, potentially reducing the device length by an order of magnitude.
Even with a moderate n g = 20, there is already a >4 times
enhancement. An added advantage of PCWs is that, unlike
rib-waveguides, they do not require the vertical partial etching
of the silicon slab which can be difficult to control and
maintain uniformity across an entire wafer. Although there
have been earlier reports on modulators incorporating PCW,
they have been capable only of simple sinusoidal modulation,
and furthermore only at speeds below 1.6 GHz [7-11].
Recently we reported the first 10 Gb/s non-return-to-zero
(NRZ) pseudo-random bit sequence (PRBS) data modulation
in a PCW-MZI modulator, and furthermore with a PCW phaseshifter length of only 200 μm [12]. This is the same length as
the shortest 10 Gb/s injection-type rib-waveguide modulator
at the time [13]. The PCW-MZI operated predominantly by
carrier-injection, and used pre-emphasis [13, 14] to achieve
10 Gb/s modulation.
Following our initial report, in this paper we demonstrate both carrier-injection- and depletion-mode modulation
at 10 Gb/s, in a 200 μm-long PCW-MZI. We perform a
more detailed analysis of injection-mode modulation with preemphasis. We also demonstrate the first depletion-mode modulation in a 200 μm MZI device, and find that pre-emphasis is
not necessary but can be beneficial at moderate strengths. In
Section II we describe the device details and perform a simple
estimate of the relation between slow-light and the device
length. In Section III we describe the experimental setup. In
Section IV we perform basic electrical, optical and electrooptic characterization of the device. Section V covers 10 Gb/s
modulation without pre-emphasis, demonstrating reverse-bias
modulation in our PCW-MZI. Section VI covers modulation
with pre-emphasis, showing that moderate pre-emphasis can
be beneficial even for reverse-bias modulation. Finally a
conclusion is made in Section VII.
II. D EVICE P RINCIPLE
Fig. 1(a) shows the schematic of our PCW-MZI modulator,
which consists of an asymmetric MZI with a p-n doped, SiO2 clad PCW [15] phase-shifter in each arm. Detailed parameters
of the device are described further below.
0018–9197/$31.00 © 2012 IEEE
NGUYEN et al.: PHOTONIC CRYSTAL SILICON OPTICAL MODULATORS: CARRIER-INJECTION AND DEPLETION AT 10 Gb/s
Metal
Doped region:
p+
p
Fast light (small ng)
ut
Signal O
n
211
ωf
k (small)
n
ω
n+
CW in
k (large)
ωs
Slow light (Large ng)
k
(a)
200 μm
PCW
(a)
104
n+
Si
wire
1 μm
n
Lπ [μm]
p
di
n( ×10−4)
−1
103
PCW
p+
−2
−4
102
−10
d+
Si Wire
101
(b)
Fig. 1. (a) Schematic of PCW-MZI modulator. (b) SEM image of a typical
CMOS-fabricated PCW with the SiO2 cladding removed. The SEM is overlaid
with a schematic of the p-n doping regions.
The device is operated by driving either (single-ended) or
both (push-pull) of the PCWs with a RF signal. As a voltage
is applied across the diode-embedded PCW, carrier movement
(injection or depletion) occurs, thereby inducing a refractive
index change n through the carrier-plasma effect [16]. This
results in an optical phase-shift ϕ in the PCW.
Fig. 2(a) shows a schematic band diagram of the PCW
mode, in terms of frequency ω and wavenumber k. The band
exhibits both fast- and slow-light regimes, where in the latter
the band becomes horizontal and n g increases towards the
band-edge. When n is applied to the PCW, the band of
the guided mode shifts vertically. This results in a shift in
wavenumber k when operating at a fixed frequency, which
is larger in the slow-light regime as shown in Fig. 2(a). This
can be quantified as k = (n g /c)ω. By expressing ω =
(∂ω/∂n)n, we see that ϕ = k L = n g (∂ω/∂n)n L/c.
Therefore a larger n g can reduce the n and/or L necessary
to obtain the required phase-shift, in other words making the
device more efficient.
Fig. 2(b) shows L π — the phase-shifter length required for
π phase shift — as a function of n g and an applied n of 10−4
order, which is reasonable by the carrier-plasma effect. Here
we estimate (∂ω/∂n) from the shift in band-edge frequency
when the material index of the PCW is increased, calculated
using 3D finite-difference time-domain simulations. We find
that for a generous n = −4 × 10−4 , Si-wire devices require
large device lengths with L π >2 mm due to their small n g<5.
On the other hand, L π can be reduced to ∼500 μm and
0
20
40
60
80
100
ng
(b)
Fig. 2. (a) Schematic band diagram of the PCW mode and (b) calculated
L π as a function of n g and n.
∼200 μm at n g = 20 and 50 respectively, both of which
can be achieved easily in PCWs. Therefore the slow-light in
PCWs can enable phase-shifter lengths that are an order-ofmagnitude shorter than in rib-waveguide devices.
Here we use PCW-MZI devices with a PCW length of
200 μm and a MZI length asymmetry of 120.1 μm. The
PCW consists of a Wl waveguide surrounded by a triangular
lattice of holes with a target diameter 2r = 215 nm and pitch
a = 400 nm. The PCW is formed in a Si slab of 220 nm
thickness, and is covered above and below by SiO2 . The
p-n region is defined by moderately-doped p (1 × 1013 cm−2 )
and n (6 × 1012 cm−2 ), and highly-doped p+ and n + (both
4 × 10 cm−2 ) regions. The device is fabricated using CMOScompatible process. Fig. 1(b) shows an SEM of a typical
PCW, with the SiO2 cladding removed, and also shows the
p-n doping regions schematically. The separations between the
p and n, and p+ and n + regions are defined as di and d+ ,
respectively.
III. E XPERIMENTAL S ETUP
Optical modulation is performed in single-ended mode by
driving the p-n junction in one of the PCW-MZI arms, by
231 −1 bit NRZ PRBS signals. The electrical driving signal,
produced by a combination of an electrical synthesizer and a
pulse pattern generator (PPG), is amplified and combined with
a DC bias through a bias-tee. This electrical signal drives the
p-n junction, modulating the TE-polarized light from a tunable
CW laser that is coupled onto the chip via a lensed fiber.
212
IEEE JOURNAL OF QUANTUM ELECTRONICS, VOL. 48, NO. 2, FEBRUARY 2012
40
I [mA]
(di , d+) [μm]
30
(0, 4)
(1.5, 4)
(3, 4)
20
(0, 6)
(0, 8)
10
0
0
1
2
VDC [V]
(a)
3
4
40
2r/a 0.54
30
r δ wmin
I [mA]
0.56
a
20
0.59
10
0.61
0
0
0.5
10
1.5
VDC [V]
2.0
2.5
3.0
(b)
Fig. 3. I-V curves of the PCW phase-shifter for (a) varying di and d+ and
(b) varying 2r/a when di = μm and d+ = 4 μm.
The output optical signal is amplified by an erbium-doped
fiber amplifier, then passed through an O/E converter. Finally
the electrical signal is detected on an Agilent 86100C/54754A
sampling oscilloscope and 18 GHz detector set, as well as a
bit-error rate (BER) tester. An Alnair Labs ORC-400 40 Gb/s
receiver and an Agilent 11982A 11 GHz receiver are used for
O/E conversion in the frequency response and data modulation
experiments, respectively. A PPG and BER tester combination
of Anritsu MU181020A/MU181040A were used for data
modulation experiments without pre-emphasis, while those in
Alnair Labs SeBERT-1040C were used in the experiments with
pre-emphasis. Lastly the electrical power of the drive signal
is measured using an electrical power meter.
IV. BASIC C HARACTERIZATION
A. Electrical Properties
Fig. 3(a) shows the I-V curves of the p-n junction embedded
in the PCWs with a target 2r = 215 nm and a range of di and
d+ . When d+ = 4 μm and di is varied from 0 to 3 μm, the
I-V curve changes only slightly, with the forward resistance R
(slope of the curve) increasing by 11% from 29 to 32 .
On the other hand, when di = 0 μm and d+ is increased from
4 μm to 8 μm, the I-V curve becomes significantly shallower,
and R increases by a factor of 2.7 from 29 to 79 . These
results indicate that a smaller d+ is preferred, although it must
also be sufficiently large to avoid excessive optical loss.
Fig. 3(b) compares the I-V curves for PCWs with target
2r = 215 − 245 nm (normalized hole-diameter 2r/a =
0.54 − 0.61), while di = 0 μm and d+ = 4 μm are
fixed. As 2r/a increases, the effective cross-section for carrier
transport decreases, causing R to increase super-linearly from
29 to 120 . While one may be tempted to reduce 2r
to lower the resistance, we note that this will also affect
strongly the optical properties of the PCW. Hence this is an
optimization problem involving 2r, a and the slab thickness,
to maximize the electrical cross-section while simultaneously
maintaining the desired PCW transmission bandwidth and
slow-light properties.
We note that in Fig. 3(b), the increase in R by a factor of
4.2 is surprisingly large, given that 2r/a is increased merely
by 14%. We believe that this is due to the depletion of carriers
around the PCW holes, caused by surface damage and dopantdeactivation during the dry-etching process to form the PCW
holes. This increases the electrical resistance across the PCW,
increasing its effective filling-fraction [17]. From a simple
analysis we estimate the depletion width δ = 73 nm, which
then explains the large increase in R [12].
An important observation here is the low forward-resistance,
despite the fact that this is a photonic-crystal-based device.
R = 29 in the device used for experiments below, and this
is 41% lower than the rib-waveguide device in Ref [13], noting
that our on-chip loss during modulation is also lower. This
indicates that the hole-matrix of the PCW does not impede the
motion of carriers in and out of the waveguiding region, any
more than the thin, partially-etched slab of the rib structure.
While care is needed in the PCW design, in terms of operating
bandwidth and the depletion zone mentioned above, PCWs
should exhibit comparable, if not superior, electrical properties
compared to rib-waveguides.
B. Optical Transmission
The PCW-MZI is integrated on-chip with low-loss spot-size
converters, resulting in an optical insertion loss as low as 13 dB
fiber-to-fiber, which is better than the previously-reported
PCW-MZI modulators [8, 9]. The insertion loss includes 6 dB
coupling loss through the spot-size converters, which can be
reduced further by improving the lithographic processes to
sharpen the Si waveguide taper. The 7 dB on-chip loss includes
∼2 dB from the MZI and the remainder from coupling into and
out of the PCW, both of which can be reduced by improved
fabrication. Furthermore, for the devices with di = 0 − 3 μm
and d+ = 4 − 8 μm, we observe no definitive increase in
optical loss. Therefore when d+ = 4 μm, the p+ and n +
regions are sufficiently separated from the center of the PCW
so as to not cause additional optical losses, at least in the
fast-light regime.
The experiments reported hereafter are performed on PCWMZI devices in which di = 0 μm and d+ = 4 μm, and using
the slow-light regime. The device used in the experiments in
the remainder of this section has R = 29 and n g ≈ 18
as with the case in Ref. [12], while the device in the data
0
−10
−20
Transmission
[dB]
Transmission [dB]
NGUYEN et al.: PHOTONIC CRYSTAL SILICON OPTICAL MODULATORS: CARRIER-INJECTION AND DEPLETION AT 10 Gb/s
VDC [V]
+0.8 −1
+0.7 −2
0
−3
0
−10
−20
−30
545
1545
1546
1547
1548
1546
1549
1547
1548
1549
1550
λ [nm]
λ [nm]
(a)
0
−10
0.7 V
−20
Transmission [dB]
0V
1.0V
Tra nsmission
[10 dB/div]
213
1.1V
1.3V
1.5V
λ = 1548.20 nm
0
λ = 1547.45 nm
−10
−20
0
λ = 1548.55 nm
−10
−20
1545
1546
1547
1548
λ [nm]
1549
1550
0
0.5
1.0
1.5
2.0
VDC [V]
(b)
(c)
Fig. 4. Optical transmission spectra (a) at VDC = −6 to +0.9 V and (b) under forward bias. (c) Bias-dependence of the optical transmission at fixed
wavelengths.
modulation experiments in Sections V and VI has R = 40 and n g ≈ 28. The devices have an estimated capacitance
(limited by measurement apparatus) of <800 fF, and hence
a high-frequency RC cutoff of >7 GHz.
C. DC Electro-Optic Response
Fig. 4(a) shows the optical transmission spectra through
a PCW-MZI at different DC bias voltages (VDC ), where
each spectrum exhibits λ-dependent oscillations from the MZI
asymmetry. As VDC is increased from −3.0 V to +0.8 V,
the spectrum tilts locally at each transmission peak, where the
transmission at the short- and long-wavelength sides of the
peaks rise and fall respectively, by up to 14 dB. This suggests
that we can achieve efficient modulation by operating at a λ
where the loss is low and the change in transmission is large.
Here we measure the transmission spectra only for VDC <
0.8 V because at stronger forward-bias, n from the carrierplasma effect becomes masked by thermal effects.
Fig. 4(b) shows the evolution of the transmission spectrum
at VDC up to +1.5 V. While the spectrum initially blue-shifts
by ∼0.1 nm at VDC = +0.7 V due to the carrier-plasma effect,
the thermo-optic effect becomes dominant at higher VDC and
the spectrum red-shifts by ∼1.4 nm at VDC = +1.5 V. The
point at which the two effects are equal can be determined
by measuring the transmission at fixed λ as a function of the
applied bias.
Fig. 4(c) shows the transmission spectrum and the biasdependent transmission at fixed λ at and on either sides of the
local transmission peak of 1548.20 nm. At λ = 1548.20 nm
the transmission initially drops very gradually until reaching
a local minimum at VDC = +0.9 V that is 1.1 dB deep.
This is due to the spectrum slightly blue-shifting from the
carrier-plasma effect. At VDC > 0.9 V the transmission
oscillates more rapidly and a larger extinction ratio (ER) is
observed, because of the larger red-shift from the thermooptic effect. For the same reason, at λ = 1547.45 nm and
1548.55 nm, the initial rise/fall in transmission is gradual but
the oscillation becomes more rapid and ER increases at higher
voltages. We can infer from the reversal of the transmission
change that the carrier-plasma and the thermo-optic effects
are balanced at VDC = 0.9 V, and the latter becomes dominant
at stronger forward-bias. Consequently we cannot infer Vπ
from the result in Fig. 4(c).
D. Vπ Measurement
We measure the modulation efficiency of the device, Vπ ,
under AC conditions so that it is not obscured by slow
thermal effects. We modulate the optical signal with a 10 MHz
sinusoidal electrical signal, and increase the peak-to-peak
voltage of the drive signal (Vpp ) until the optical peaks/troughs
begin to overturn. Fig. 5 show the oscilloscope traces of the
modulated optical signal at zero- and forward-bias voltages.
Without a DC bias, we measure Vπ = 2.84 V at λ =
1546.98 nm. It is a relatively large value, however we note
that the drive signal waveform varies from −1.4 V to +1.4 V,
but the negative voltage part results in little phase-shift. In this
case the peaks of the optical signal correspond to a negative
drive voltage, hence becoming flat-topped - the optical signal
is vertically asymmetric about zero. On the other hand, when
the device is forward-biased to VDC = +1.1 V, we measure
Vπ = 0.8 V at λ = 1547.23 nm and the optical signal becomes
more sinusoidal and vertically symmetric, because the drive
214
IEEE JOURNAL OF QUANTUM ELECTRONICS, VOL. 48, NO. 2, FEBRUARY 2012
60
VDC 0 V
Vπ 2.8 V
0
50
0
VDC 1.1 V
Vπ 0.8 V
0
−10
40
30
−20
ng
−1
1
Transmission [dB]
Optical signal [a.u.]
1
20
−30
10
−1
Time [50 ns/div]
Fig. 5. Sampling oscilloscope traces of the optical signal modulated with a
10 MHz RF signal, at VDC = 0 V and 1.1 V.
Norm. modulation depth [dB]
−2
−6
−8
106
Fig. 6.
1545
1550
0
1555
λ [nm]
Fig. 7. Optical transmission and n g spectra of the PCW-MZI used in the
data modulation experiments.
0
−4
−40
1540
VDC
[V]
Vpp
[V]
f3dB
[GHz]
0.8
0.024
1
0
0.25
3
−3
0.77
6
107
108
109
Frequency [Hz]
1010
1011
Small-signal frequency response of the PCW-MZI at different VDC .
voltage is now completely in the carrier-injection regime, from
+0.7 V to +1.5 V. These values of Vπ correspond to a figure
of merit Vπ L = 0.056 V · cm and 0.016 V · cm, which are
comparable to or smaller than other carrier-injection type MZI
modulators [13]. We note again that these results arise from
the low forward-resistance in our PCW-based device.
In the case of Vπ under reverse-bias, it can be measured
under DC conditions since there is no carrier-injection nor the
subsequent thermal effects involved. However we were unable
to reach Vπ due to our short device length, hence we can only
conclude that, under reverse-bias, Vπ L > 0.36 V · cm.
E. Frequency Response
We characterize the small-signal frequency response of the
PCW-MZI by observing the modulation depth of the optical
signal as a function of the drive frequency. Here the device
is driven by a sinusoidal signal from 50 MHz to 15 GHz,
and the modulated optical signal is optically amplified, O/E
converted using the 40 Gb/s receiver, and viewed on the
sampling oscilloscope.
Fig. 6 shows the frequency response measured at different
VDC . To compare the response at different bias, the Vpp and
λ are adjusted slightly as indicated in Fig. 6, such that the
modulation depth at 50 MHz is the same for each VDC .
The measured response are then normalized to their values
at 50 MHz. We find that the 3 dB bandwidth ( f 3dB ) is
approximately 3 GHz when no bias is applied (VDC = 0 V).
However, ( f 3dB ) decreases to 1 GHz under forward-bias
(VDC = +0.8 V), because here the modulation occurs predominantly via carrier-injection and involves carrier-diffusion
which is slow. On the other hand f 3dB increases to 6 GHz
under reverse-bias (VDC = −3 V), since modulation occurs
via carrier-depletion and carriers are moved in and out of
the PCW more rapidly. More investigation into the device
capacitance is needed, in order to determine whether the 6
GHz bandwidth is RC-limited. Furthermore, the frequency
response may have been affected by the roll-off of the 12.5
GHz electrical amplifier, as well as the relatively low signal-tonoise ratio and weak signal strength of the optical amplifier
and O/E converter combination. An improved measurement
setup may reveal a larger bandwidth.
Although f 3dB will vary with Vpp particularly under
forward-bias, the results in Fig. 6 indicate that for the same
low-frequency modulation depth, the device exhibits a larger
bandwidth under reverse-bias.
V. 10 Gb/s M ODULATION WITHOUT P RE -E MPHASIS
A. Device Properties
For the modulation experiments hereafter, we use a device
that is different from, but has the same design parameters
as, the device used for the experiments in Section IV. The
modulation characteristics of the earlier device is reported in
Ref. [12]. The forward resistance of this device is 40 which
is 38% larger than the earlier device.
Fig. 7 shows the optical transmission and n g spectra of the
PCW-MZI device used hereafter. The transmission spectrum
is normalized to the peak transmission, and exhibits oscillations due to the asymmetric nature of the MZI. The n g
spectrum is obtained from group delay measurements using
the modulation phase shift method [18]. This method gives
inaccurate results around the transmission dips in Fig. 7
where the MZI arms interfere destructively, hence n g at these
wavelengths are not shown. As λ increases towards the bandedge, n g increases, as expected.
76.9 mV/div
NGUYEN et al.: PHOTONIC CRYSTAL SILICON OPTICAL MODULATORS: CARRIER-INJECTION AND DEPLETION AT 10 Gb/s
2 Gb/s
Vpp = 2.9 V
Loss = 0.84 dB
ER = 6.4 dB
76.9 mV/div
[100 ps/div]
(a)
5 Gb/s
Vpp = 2.9 V
5 Gb/s
Vpp = 2.0 V
Loss = 4.1 dB
ER = 7.5 dB
24.2 mV/div
76.9 mV/div
[50 ps/div]
(b)
10 Gb/s
Vpp = 1.9 V
Loss = 6.8 dB
ER = 3.6 dB
10.2 mV/div
76.9 mV/div
[50 ps/div]
(c)
[20 ps/div]
(d)
Fig. 8. Eye patterns of NRZ PRBS signals at (a) 2 Gb/s and (b) 5 Gb/s,
with the same Vpp = 2.9 V. Eye pattern at (c) 5 Gb/s and (d) 10 Gb/s, each
with Vpp reduced until the eye visually appears most open.
Both the transmission and n g spectra become noisy at long
λ. This is because a slight shift in the PCW bandwidth between
the MZI arms, due to fabrication imperfections, results in a
large and random phase-shift particularly in the slow-light
regime, and manifest as noise in the spectra. We perform our
modulation experiments in the slow-light regime as before,
between the transmission peak and dip at λ = 1548.0 nm and
1548.7 nm, respectively, where n g ≈ 28, FSR ≈ 1.3 nm and
ER (between the local transmission peak and dip) of ∼17 dB.
B. Zero-Bias Operation
Fig. 8 shows the eye patterns for when the device is driven
by 231−1 bit NRZ PRBS signals at 2 Gb/s, 5 Gb/s and 10 Gb/s.
215
The modulation is performed without pre-emphasis, with zero
bias (VDC = 0 V) and drive signal Vpp = 1.9 − 2.9 V as
indicated on each diagram, while the operating wavelength
is fixed at λ = 1548.60 nm. Fig. 8(a) shows that the eye is
open at 2 Gb/s with ER = 6.4 dB, which is taken to be the
ratio of the difference in the 1- and 0-levels to the difference
in the 1-level and the noise-level, each measured on the
sampling oscilloscope. The noise-level corresponds to when
the input to the optical amplifier preceding the O/E converter
is disconnected. The in-device loss, which we define as the
loss of the 1-level signal strength relative to the maximum
transmission through the device, is 0.84 dB.
We find that the rise-time (τ rise ) of the eye is shorter than
the fall-time (τ fall ), while the timing jitter is larger. This can
be explained as follows. In our experiments, when we consider only the carrier-plasma effect the transmission decreases
(increases) when a positive (negative) voltage is applied. This
is because our operating wavelength is on the long-wavelength
side of the local transmission peak and that we are driving the
PCW on the longer arm of the asymmetric MZI. Therefore the
0- and 1-levels of the eye pattern correspond to the carrierinjected and depleted states, respectively. The 1-0 transition
involves the injection of carriers through diffusion and is slow,
resulting in a long τ fall . On the other hand, the 0-1 transition
involves a more rapid extraction of carriers through the applied
reverse potential and thus a shorter τ rise , however a large
timing jitter arise depending on the amount of carriers injected
in the preceding bit pattern.
When the bitrate is increased to 5 Gb/s while keeping V pp
constant, the eye closes due to the long τ fall , as seen in
Fig. 8(b). The signal quality can be improved by reducing
Vpp and consequently the amount of injected carriers, but this
affects both the modulation strength and the loss as shown in
Fig. 8(c,d), in which case it is questionable whether the eye
is “open”.
We can overcome to a certain extent the bandwidth limitations due to the slow carrier diffusion, in one of two ways. One
way is to reverse-bias the device such that it operates mostly,
if not completely, by carrier-depletion. The other method is to
pre-emphasize the drive signal [13, 14]. These are described
below.
C. Reverse-Bias Operation
Fig. 9(a) shows the 10 Gb/s eye pattern under a reverse-bias
of VDC = −3.0 V, at λ = 1548.74 nm. The drive signal has
Vpp = 5.9 V such that the device is operating completely in
carrier-depletion mode. The eye is clearly open, even at 10
Gb/s. Here, τrise = 89 ps and τfall = 96 ps are roughly equal
because of completely reverse-bias operation, and the rootmean-square (RMS) jitter is 8.7 ps. The rise (fall) of the eye
exhibit “double lines” that represent the distinct trajectories
of the 101 (010) and 001 (110) bit sequences. We show in
Section VI that this can be removed by moderate pre-emphasis
of the drive signal. However, even without pre-emphasis we
measure BER = 2 × 10−9 at a received optical power of
−2.5 dBm. We also measure an in-device loss of 1.9 dB (total
on-chip loss of 8.9 dB) and ER = 4.0 dB. The electrical power
216
IEEE JOURNAL OF QUANTUM ELECTRONICS, VOL. 48, NO. 2, FEBRUARY 2012
VDC = −3.0 V
VPP = 5.9 V
BER = 2×10−9 Loss = 1.9 dB
Vpp = 3.8 V
Overshoot = 42%
[20 ps/div]
[20 ps/div]
Amplitude
=1.9 V
(a)
ER = 4.0 dB
Loss = 1.8 dB
BER = 2 × 10
−4
BER
Drive voltage
[V]
(a)
2
0
−2
−4
−6
10−2
10−4
VDC −2.0 V
10−6
[20 ps/div] ER = 5.8 dB
VDC −3.0 V
(b)
10−8
Fig. 10.
Eye patterns of (a) 10 Gb/s pre-emphasized drive signal and
(b) corresponding modulated optical signal.
0
−5
−10
10
ER
[dB]
Transmission
[dB]
10−10
5
0
4
5
6
Drive Vpp [V]
7
(b)
Fig. 9.
(a) Eye pattern under reverse-bias, without pre-emphasis.
(b) Evolution of the drive voltage, BER, in-device transmission and ER, as a
function of Vpp of the drive signal.
of the drive signal is measured to be 9.6 mW, corresponding
to a RF power consumption of 9.6 pJ/bit.
For a given λ and VDC the eye can be optimized by
varying Vpp of the drive signal. Fig. 9(b) shows the voltage
range of the drive signal, as well as the BER, in-device
transmission (negative of the in-device loss) and ER of the
modulated optical signal, as a function of Vpp . These are
measured at the same, fixed λ as in Fig. 9(a). The blue and
green curves represent data for VDC = −3.0 V and −2.0 V,
respectively.
For VDC = −3.0 V and Vpp in the range of 4 V to 7 V, we
find that the BER is minimum at Vpp = 5.9 V, for which the
eye pattern is shown in Fig. 9(a). On the other hand the indevice transmission increases from −6 dB to 0 dB, while the
ER remains roughly constant at 4 dB to 5 dB. Out of these,
the result for Vpp = 5.9 V appears most favorable, taking into
account the low BER of 10−9 order, relatively low loss of
1.9 dB and an ER of 4.0 dB. We note that this occurs when
Vpp ∼ 2|VDC |, because at larger Vpp the drive signal begins
to swing into a positive voltage as shown in Fig.9(b), where
carrier-injection and diffusion begin to occur and degrade the
modulated signal.
Similar observations can be made about the results for
VDC = −2.0 V, although the lowest BER is larger at 2 ×10−8 .
However, we believe the BER can be reduced further by also
optimizing λ, since the optimal λ will be different for each
VDC . Further investigations are required to determine the best
parameter combination to optimize the BER, transmission and
ER, while keeping Vpp reasonably low.
D. Summary
In summary, a reverse-bias operation is necessary for
our PCW-MZI modulator to operate at 10 Gb/s without pre-emphasis. While there have been reports of submm carrier-depletion type MZI modulators [19, 20], this
is the first demonstration of 10 Gb/s modulation in a
MZI modulator as short as 200 μm, without requiring
pre-emphasis.
We believe that depletion-mode modulation is made possible in such a short device, owing in part to the slowlight, where an n g ∼ 28 would give a ∼7 times modulation
efficiency enhancement compared to other devices using ribwaveguides. There may be other factors involved, such as the
short device length helping to reduce the loss of the electrical
drive signal; these require further investigation.
VI. 10 Gb/s M ODULATION WITH P RE -E MPHASIS
A. Zero-Bias Modulation
Fig. 10 shows the modulation results at 10 Gb/s
using pre-emphasized drive signals [13, 14], performed at
λ = 1548.79 nm. Fig. 10(a) shows the eye pattern of
NGUYEN et al.: PHOTONIC CRYSTAL SILICON OPTICAL MODULATORS: CARRIER-INJECTION AND DEPLETION AT 10 Gb/s
VDC = −0.40 V
VDC = +0.65 V
VDC = 0 V
Vpp = 5.8 V
217
Loss = 0.3 dB
BER = 3 × 10−7
[20 ps/div]
ER = 2.7 dB
(a)
Drive Vpp
[V]
4.5
Fig. 12. Eye pattern of the 10 Gb/s modulation under reverse-bias, combined
with a slight pre-emphasis. Inset shows the eye pattern of the pre-emphasized
drive signal.
4.0
3.5
device loss is 1.8 dB (total on-chip loss of 8.8 dB), ER =
5.8 dB and the RF power consumption is 2.6 pJ/bit. This
is nearly 2 orders larger than ring-type modulators, but still
is the smallest for a MZI modulator as far as we are aware
[1, 13].
3.0
BER
10−2
10−3
Transmission
[dB]
10−4
B. Bias-Dependent Modulation
0
−2
−4
ER
[dB]
10
5
0
−0.5
0
0.5
1.0
VDC [V]
(b)
Fig. 11.
(a) Eye patterns of the pre-emphasized drive signal and the
corresponding optical signals at different VDC , using pre-emphasis, at various
VDC . (b) Evolution of the drive Vpp , BER, in-device transmission and ER, as
a function of VDC .
the pre-emphasized drive signal, where the sharp rise/fall
at the beginning/end of each electrical pulse enables faster
injection/depletion of carriers. VDC = 0 V and the signal
amplitude is 1.9 V, while Vpp = 3.8 V including the preemphasis spike. We note that the eye is measured with
an 18 GHz detector, and the actual overshoot of the preemphasis may be sharper and stronger than observed in Fig.
10(a). Fig. 10(b) shows the eye pattern of the modulated
optical signal. The eye is clearly open, indicating that preemphasis has successfully compensated for the carrier-related
bandwidth limitation. The timing jitter is 8.8 ps and is
comparable to the reverse-bias case without pre-emphasis.
The amplitude noise is larger and the BER is worsened to
2 × 10−4 , but this is still below the typical threshold (10−3)
for receivers employing forward error-correction [21]. The
received power of the signal is −3.0 dBm, while the in-
Fig. 11(a) shows the evolution of the 10 Gb/s preemphasized eye pattern as a DC bias is applied. At each
VDC , the eye pattern is optimized by adjusting the drive signal
amplitude and the pre-emphasis level, while λ is fixed. Under
forward-bias, the optical eye is optimized by simultaneously
reducing the drive signal amplitude and strengthening the preemphasis. This results in an increased signal transmission,
indicated by the larger height of the eye in Fig. 11(a). On the
other hand, under a reverse-bias the optimal eye is achieved
by increasing the drive signal amplitude while weakening the
pre-emphasis. However the signal transmission decreases as
seen by the smaller eye, and the BER increases.
Fig. 11(b) summarizes the evolution of the BER, signal
transmission and ER as a function of VDC , as well as the
evolution of drive signal Vpp . In the −0.2 < VDC < 0.8 V range
the BER and ER remain roughly constant at ∼2×10−4 and 5
to 6 dB, respectively, while both the signal transmission and
drive Vpp improve. At stronger forward-bias the BER increases
because of insufficient pre-emphasis to suppress the increasing
effect of carrier diffusion. Nevertheless this suggests that a
moderate forward-bias can be beneficial, in this case increasing
the signal transmission by 1.7 dB while lowering the drive Vpp
by 10%.
While the above results seem to indicate that with preemphasis the forward-bias modulation performs superior to
reverse-bias modulation, we note that this is the case for when
λ is fixed and initially optimized for VDC = 0 V. By also
adjusting λ, it is possible to obtain a more superior eye under
reverse-bias, as we show below.
C. Reverse-Bias with Pre-Emphasis
Fig. 12 shows the 10 Gb/s eye pattern when the PCWMZI is driven under a reverse-bias of VDC = −3.0 V and
218
IEEE JOURNAL OF QUANTUM ELECTRONICS, VOL. 48, NO. 2, FEBRUARY 2012
VDC = 0 V
VPP = 1.6 V
Injection-mode
BER = 4 × 10−4
Depletion-mode
10
8
(a)
VDC = −3 V
VPP = 4.0 V
BER = 8 × 10−6
IBM (2007)
Sandia N.L.
(2010)
6
Vpp [V]
Loss = 2.3 dB
[20 ps/div] ER = 4.6 dB
Intel (2007)
Surrey (2011)
Our work
U. Paris-sud
(2011)
IME (2010)
4
2
UCSB
(2010)
Bell Labs
(2010)
Intel (2005)
Fujitsu (2010)
0
10−4
Loss = 0.3 dB
[20 ps/div] ER = 3.0 dB
10−3
Phase-shifter length [m]
10−2
(b)
Fig. 13. Eye patterns of the pre-emphasized 10 Gb/s modulation without
RF termination, in (a) injection mode and (b) depletion mode. Insets show
the eye pattern of the drive signals.
by a moderately pre-emphasized signal with Vpp = 5.6 V
(shown in the inset), at λ = 1548.73 nm. We find that the
“double lines” that were visible in the eye pattern without
pre-emphasis in Fig. 9(a) are no longer present, giving rise
to a cleaner eye pattern. Here we measure an in-device loss
of 0.3 dB (total on-chip loss of 7.3 dB) and ER = 2.7 dB,
which are comparable to the case without pre-emphasis, as
well as a reduced RMS timing jitter of 5.3 ps. The RF power
consumption is 7.9 pJ/bit. The BER = 4.7 × 10−7 is larger
compared to the non-pre-emphasized case; however this may
be due to the different error-detector used. In any case we
believe that the signal quality can be further improved by a
more detailed optimization of the modulation parameters, and
that a moderate pre-emphasis improves the signal quality of
reverse-bias modulation.
D. Modulation without RF Termination
All of the RF modulation results thus far have been performed with the back-end of the device connected to a RF
probe attached to a 50 terminator. By removing this probe
and leaving the device electrically open, we find that the drive
voltage of the modulator can be reduced further.
Fig. 13 shows the 10 Gb/s eye pattern in injection- and
depletion-modes without the back-end RF probe, with both
eyes open. Compare to the case with the 50 termination,
Vpp is lowered by 58% to 1.6 V in injection-mode, and by
31% to 4.0 V in depletion-mode. Visually, additional noise in
the eye diagrams is barely noticeable. The BER is increased,
but in injection-mode it is still of the same order at 4 × 10−4 ,
while in depletion-mode it is increased by nearly 2 orders-ofmagnitude to 8 × 10−6 . Although not shown here, similarly
in depletion-mode without pre-emphasis as in Section V, Vpp
Fig. 14. Comparison of our results with other Si MZI modulators that operate
at >10 Gb/s, in terms of the drive voltage and the phase-shifter length.
is reduced by 44% to 3.6 V and BER increases to 2×10−8 .
Nevertheless the BERs are still below the typical threshold
for FEC-employed receivers, and in some cases the benefit of
reduced Vpp may outweigh the increased BER. The mechanism
behind the reduced Vpp in the absence of the 50 termination
is a matter of further investigation.
E. Summary
In summary, pre-emphasis can enable 10 Gb/s modulation
even when carrier-injection occurs, such as in the case of zerobias operation. Generally a stronger pre-emphasis is required
when driving the device under forward bias. Even when
driving the device completely in depletion-mode, a moderate
pre-emphasis gives rise to a cleaner signal. In addition, device
operation without the 50 RF termination can reduce the
Vpp further, at a slight expense of the BER. As mentioned in
Section V, we believe that slow-light (n g ≈ 28) is playing a
major role in making depletion-mode modulation possible in
a device of only 200 μm length.
VII. C ONCLUSION
We have demonstrated 10 Gb/s modulation in a 200 μm
photonic crystal silicon optical modulator in both carrierinjection and depletion modes, with the drive voltages as
low as 1.6 V and 3.4 V, respectively. In injection-mode, preemphasis was necessary to compensate for the carrier-related
bandwidth limitation. In depletion-mode, pre-emphasis was
not required, although at moderate levels it improved the signal
quality.
Fig. 14 summarizes our results in comparison with other
Si MZI modulators that operate at >10Gb/s [2-4, 13, 19, 20,
22-25], in terms of the device length and the drive voltage.
In particular, this is the first depletion-mode operation in a
NGUYEN et al.: PHOTONIC CRYSTAL SILICON OPTICAL MODULATORS: CARRIER-INJECTION AND DEPLETION AT 10 Gb/s
MZI modulator this short, where only a few others have submillimeter lengths [19, 20, 25].
We believe that the depletion-mode modulation in a short,
200 μm device was possible owing to the slow-light enhancement, with n g ∼28. Further investigations are required to
determine whether there are other contributing factors that
has made this possible in such a short device. There may be,
for example, improved phase-matching between the electrical
drive signal and the slow-light. In that case, a larger n g may
not necessarily be beneficial, but rather there may be a range
of n g that optimizes both the signal–slow-light phase-matching
and the n g -enhancement of ϕ.
While Fig. 14 is not the complete story as there are other
factors such as loss, extinction ratio and operating bandwidth,
the comparison indicates nevertheless the potential for future
compact, low-voltage Si MZI modulators. As for the operating
bandwidth of PCW modulators, in general it becomes smaller
at larger n g due to the larger second-order dispersion. It is
possible, however, to maximize the bandwidth by dispersion
engineering the PCW [26]. In the ideal case, we assume that
the slow-light band in the band-diagram is straight (has a
constant slope) between the light-line and the Brillouin zone
edge. In this case the bandwidth can be approximated by the
relation n g ( f / f ) ∼ n m , where ( f / f ) is the normalized
bandwidth, and n m is the change in modal index within
the bandwidth. For air-clad PCWs n m ≈ 0.6 [26], while
this is roughly halved for SiO2 -clad PCWs used in this work.
Hence with n g = 30 and at communications wavelengths,
we can expect air- and SiO2 -clad PCW modulators to have
a bandwidth of ∼31 nm and ∼16 nm, respectively.
We anticipate further improvements in the device performance, for example by optimizing the p-n doping profile
to the slow-light mode profile to increase the modulation
efficiency. The on-chip losses mentioned above include ∼5
dB from coupling in/out of the PCW and ∼2 dB from other
structures, hence we anticipate these numbers to be reduced
with improved fabrication. Furthermore, such a short device
may possibly remove the need for complex travelling wave
electrodes [2, 25], while the fact that it is PCW-based removes
the need for partial-etching of the silicon slab associated with
rib-structures.
R EFERENCES
[1] G. T. Reed, G. Mashanovich, F. Y. Gardes, and D. J. Thomson, “Silicon
optical modulators,” Nature Photon., vol. 4, pp. 518–526, Aug. 2010.
[2] A. S. Liu, L. Liao, D. Rubin, H. Nguyen, B. Ciftcioglu, Y. Chetrit,
N. Izhaky, and M. Paniccia, “High-speed optical modulation based on
carrier depletion in a silicon waveguide,” Opt. Exp., vol. 15, no. 2, pp.
660–668, Jan. 2007.
[3] G. Rasigade, M. Ziebell, D. Marris-Morini, J. M. Fedeli, F. Milesi, P.
Grosse, D. Bouville, E. Cassan, and L. Vivien, “High extinction ratio
10 Gbit/s silicon optical modulator,” Opt. Exp., vol. 19, no. 7, pp. 5827–
5832, Mar. 2011.
[4] T. Y. Liow, K. W. Ang, Q. Fang, J. F. Song, Y. Z. Xiong, M. B. Yu, G.
Q. Lo, and D. L. Kwong, “Silicon modulators and germanium photodetectors on SOI: Monolithic integration, compatibility, and performance
optimization,” IEEE J. Sel. Topics Quantum Electron., vol. 16, no. 1,
pp. 307–315, Jan.–Feb. 2010.
[5] G. Rasigade, D. Marris-Morini, L. Vivien, and E. Cassan, “Performance
evolutions of carrier depletion silicon optical modulators: From p-n to
p-i-p-i-n diodes,” IEEE J. Sel. Topics Quantum Electron., vol. 16, no. 1,
pp. 179–184, Jan.–Feb. 2010.
219
[6] S. J. Spector, C. M. Sorace, M. W. Geis, M. E. Grein, J. U. Yoon, T. M.
Lyszczarz, E. P. Ippen, and F. X. Kartner, “Operation and optimization
of silicon-diode-based optical modulators,” IEEE J. Sel. Topics Quantum
Electron., vol. 16, no. 1, pp. 165–172, Jan.–Feb. 2010.
[7] T. Tanabe, K. Nishiguchi, E. Kuramochi, and M. Notomi, “Low power
and fast electro-optic silicon modulator with lateral p-i-n embedded
photonic crystal nanocavity,” Opt. Exp., vol. 17, no. 25, pp. 22505–
22513, Dec. 2009.
[8] X. N. Chen, Y. S. Chen, Y. Zhao, W. Jiang, and R. T. Chen, “Capacitorembedded 0.54 pJ/bit silicon-slot photonic crystal waveguide modulator,” Opt. Lett., vol. 34, no. 5, pp. 602–604, Mar. 2009.
[9] X. L. Wang, C. Y. Lin, S. Chakravarty, J. D. Luo, A. K. Y. Jen, and R. T.
Chen, “Effective in-device r(33) of 735 pm/V on electro-optic polymer
infiltrated silicon photonic crystal slot waveguides,” Opt. Lett., vol. 36,
no. 6, pp. 882–884, Mar. 2011.
[10] G. Shambat, B. Ellis, M. A. Mayer, A. Majumdar, E. E. Haller, and
J. Vuèkoviæ, “Ultralow power fiber-coupled gallium arsenide photonic
crystal cavity electro-optic modulator,” Opt. Exp., vol. 19, no. 8, pp.
7530–7536, 2011.
[11] C. Y. Lin, X. L. Wang, S. Chakravarty, B. S. Lee, W. C. Lai, J. D.
Luo, A. K. Y. Jen, and R. T. Chen, “Electro-optic polymer infiltrated
silicon photonic crystal slot waveguide modulator with 23 dB slow light
enhancement,” Appl. Phys. Lett., vol. 97, no. 9, pp. 093304-1–093304-3,
Aug. 2010.
[12] H. C. Nguyen, Y. Sakai, M. Shinkawa, N. Ishikura, and T. Baba,
“10 Gb/s operation of photonic crystal silicon optical modulators,” Opt.
Exp., vol. 19, no. 14, pp. 13000–13007, Jul. 2011.
[13] W. M. J. Green, M. J. Rooks, L. Sekaric, and Y. A. Vlasov, “Ultracompact, low RF power, 10 Gb/s silicon Mach-Zehnder modulator,” Opt.
Exp., vol. 15, no. 25, pp. 17106–17113, Dec. 2007.
[14] Q. F. Xu, S. Manipatruni, B. Schmidt, J. Shakya, and M. Lipson,
“12.5 Gbit/s carrier-injection-based silicon micro-ring silicon modulators,” Opt. Exp., vol. 15, no. 2, pp. 430–436, Jan. 2007.
[15] T. P. White, L. O’Faolain, J. T. Li, L. C. Andreani, and T. F.
Krauss, “Silica-embedded silicon photonic crystal waveguides,” Opt.
Exp., vol. 16, no. 21, pp. 17076–17081, Oct. 2008.
[16] R. A. Soref and B. R. Bennett, “Electrooptical effects in silicon,” IEEE
J. Quantum Electron., vol. 23, no. 1, pp. 123–129, Jan. 1987.
[17] A. Berrier, M. Mulot, G. Malm, M. Östling, and S. Anand, “Carrier
transport through a dry-etched InP-based 2-D photonic crystal,” J. Appl.
Phys., vol. 101, no. 12, pp. 123101-1–123101-6, Jun. 2007.
[18] D. Mori, S. Kubo, H. Sasaki, and T. Baba, “Experimental demonstration
of wideband dispersion-compensated slow light by a chirped photonic
crystal directional coupler,” Opt. Exp., vol. 15, no. 9, pp. 5264–5270,
Apr. 2007.
[19] M. R. Watts, W. A. Zortman, D. C. Trotter, R. W. Young, and A. L.
Lentine, “Low-voltage, compact, depletion-mode, silicon Mach-Zehnder
modulator,” IEEE J. Sel. Topics Quantum Electron., vol. 16, no. 1, pp.
159–164, Jan.–Feb. 2010.
[20] D. M. Gill, S. S. Patel, M. Rasras, K. Y. Tu, A. E. White, Y. K. Chen,
A. Pomerene, D. Carothers, R. L. Kamocsai, C. M. Hill, and J. Beattie,
“CMOS-compatible Si-ring-assisted Mach-Zehnder interferometer with
internal bandwidth equalization,” IEEE J. Sel. Topics Quantum Electron.,
vol. 16, no. 1, pp. 45–52, Jan.–Feb. 2010.
[21] E. Ip, A. P. T. Lau, D. J. F. Barros, and J. M. Kahn, “Coherent detection
in optical fiber systems,” Opt. Exp., vol. 16, no. 2, pp. 753–791, Jan.
2008.
[22] S. Akiyama, T. Kurahashi, T. Baba, N. Hatori, T. Usuki, and T.
Yamamoto, “A 1 V peak-to-peak driven 10-Gb/s slow-light silicon
Mach–Zehnder modulator using cascaded ring resonators,” Appl. Phys.
Exp., vol. 3, no. 7, pp. 072202-1–072202-3, 2010.
[23] L. Liao, D. Samara-Rubio, M. Morse, A. S. Liu, D. Hodge, D. Rubin, U.
D. Keil, and T. Franck, “High speed silicon Mach-Zehnder modulator,”
Opt. Exp., vol. 13, no. 8, pp. 3129–3135, Apr. 2005.
[24] F. Y. Gardes, D. J. Thomson, N. G. Emerson, and G. T. Reed, “40 Gb/s
silicon photonics modulator for TE and TM polarizations,” Opt. Exp.,
vol. 19, no. 12, pp. 11804–11814, Jun. 2011.
[25] H. W. Chen, Y. H. Kuo, and J. E. Bowers, “25 Gb/s hybrid silicon
switch using a capacitively loaded traveling wave electrode,” Opt. Exp.,
vol. 18, no. 2, pp. 1070–1075, Jan. 2010.
[26] T. Baba, J. Adachi, N. Ishikura, Y. Hamachi, H. Sasaki, T. Kawasaki,
and D. Mori, “Dispersion-controlled slow light in photonic crystal
waveguides,” Proc. Jpn. Acad. Ser. B: Phys. Biol. Sci., vol. 85, pp. 443–
453, Dec. 2009.
220
IEEE JOURNAL OF QUANTUM ELECTRONICS, VOL. 48, NO. 2, FEBRUARY 2012
Hong C. Nguyen received the B.Sc. (Hons) and
Ph.D. degrees in physics from the Center for
Ultrahigh-bandwidth Devices for Optical Systems
(CUDOS), School of Physics, University of Sydney,
Australia, in 2003 and 2008, respectively.
He is currently a Research Associate with the
Department of Electrical and Computer Engineering,
Yokohama National University, Yokohama, Japan.
His current research interests include nonlinear
optics and silicon photonics.
Dr. Nguyen is a member of the Optical Society of
America (OSA) and Japan Society of Applied Physics, and was the President
of the OSA Student Chapter at the University of Sydney in 2006/2007. He
received the Optium Student Prize at OECC/ACOFT in 2008 and was also a
finalist for the OSA/New Focus Bookham Student Prize in 2007.
Yuya Sakai received the B.E. degree from the
Department of Electrical and Computer Engineering,
Yokohama National University, Yokohama, Japan, in
2011. He is currently pursuing the M.D. degree in
silicon photonics optical modulator with the same
university.
He is a member of the Japan Society of Applied
Physics.
Mizuki Shinkawa received the B.E. degree from the
Department of Electrical and Computer Engineering,
Yokohama National University, Yokohama, Japan, in
2010. She is currently pursuing the M.D. degree in
nonlinear silicon photonics devices with the same
university.
She is a member of the Japan Society of Applied
Physics.
Norihiro Ishikura received the B.E. and M.E.
degrees from the Department of Electrical and Computer Engineering, Yokohama National University,
Yokohama, Japan, in 2009 and 2011, respectively.
He is currently pursuing the Ph.D. degree in silicon
photonics and photonic crystal slow light devices
with the same university.
He is a member of the Japan Society of Applied
Physics.
Toshihiko Baba (M’03) received the Ph.D. degree
from the Division of Electrical and Computer
Engineering, Yokohama National University (YNU),
Yokohama, Japan, in 1990. During his Ph.D. work,
he had been engaged in on-Si waveguides, ARROW
waveguides, and 3-D photonic integration.
He joined the Tokyo Institute of Technology,
Tokyo, Japan, as a Research Associate from 1991 to
1993. He discussed the spontaneous emission control
in vertical-cavity surface-emitting lasers (VCSELs)
and achieved the room temperature cw operation in a
long wavelength device. He became an Associate Professor and full Professor
of YNU in 1994 and 2005, respectively. In these 15 years, he has studied
photonic nanostructures such as photonic crystals (PCs), high-index-contrast
structures, and Si photonics. He first demonstrated PC waveguides, surfacePC LEDs, and Si photonic wire components. He also achieved the room
temperature cw operation in PC nanolasers and microdisk lasers with the
strong Purcell effect, record high single-mode power holey VCSEL, negative
refractive components for lightwaves, and ultra-compact Si AWG demultiplexer. His current research interests include slow light in PC waveguides
toward optical buffering and nonlinear enhancement.
Prof. Baba is a member of the Institute of Electronics, Information and
Communication Engineers, the Japan Society of Applied Physics, the Optical
Society of America, and the American the Physics Society. He received nine
academic awards including the Japan Society of Applied Physics Fellow
Award Stipulations Award in 2005, the Lasers and Electro-Optics Society
Distinguished Lecturer Award from 2006 to 2007, and the Institute of
Electronics, Information and Communication Engineers Electronics Society
Award in 2011.