COMMENTARY

High speed phase modulators for silicon photonic integrated

circuits: a role for lithium niobate?
Roel Baetsa,b,* and Bart Kuykena,b
a
Ghent University–imec, Photonics Research Group, Gent, Belgium
b
Ghent University, Center for Nano- and Biophotonics, Gent, Belgium

Lithium niobate (LN) has been and still is a prominent material for high LN has been the most successful material for high speed optical
speed optical modulators. The performance of such LN-modulators— modulation, due to its strong Pockels effect, first as a bulk material
as a stand-alone building block for optical systems—has been unchal- and later in the form of planar waveguide-based components. While
lenged for many years. But in many ways it has proven to be very hard the basic modulation effect is based on index modulation, and therefore
to integrate lithium niobate modulators with other functions on a single optical phase modulation, it is relatively straightforward to convert that
optical chip. As a result, scientific interest in this material has been into amplitude modulation, either by combining the modulator with
fading away. As a consequence of the success in the development polarizers or by making use of an interferometric construct in which
of thin film lithium niobate, recent reports1,2 are starting to change this one path is phase modulated and another path is not or inversely modu-
and may open a route to integrated photonic chips including LN-mod- lated. Upon recombination of both paths amplitude modulation results.
ulators. Furthermore these integrated LN-modulators take advantage of The key performance metric of such a modulator is the voltage V π
recent concepts (e.g., tight confinement through a high refractive index- needed for a π phase shift of the optical beam (translating into on-
contrast) and integration technologies (e.g., die-to-wafer bonding), off amplitude modulation). This voltage can be scaled down by making
allowing them to outperform their more conventional predecessors the device length L larger. Actually, the key metric is the V π Lπ product,
in sheer performance. So we can rightfully ask the question: is lithium expressed in V · cm. Typically LN-modulators have V π Lπ -product of
niobate back in business? This paper puts this question in perspective. the order of 10–20 V · cm. But making the device longer creates other
Ever since the seminal paper by Stewart Miller in 19693 there has problems. Not only is compactness an asset in its own right, but fur-
been the ambition to build integrated optical systems on a chip, very thermore the waveguide losses per unit length α (in dB/cm) yield a
much following the technical and economic logic behind the success of maximum meaningful length for the device. In this context the product
electronic integrated circuits. In such integrated optical circuits—now V π Lα (expressed in V · dB) is also often used as a figure of merit for
more commonly called photonic integrated circuits or PICs—a variety modulators. Another length-dependent property relates to the band-
of optical building blocks would be integrated on a planar substrate and width of the electro-optic conversion process. The intrinsic bandwidth
would be interconnected by means of optical waveguides. Actually the of the Pockels effect is ultra-high but the practical bandwidth is gen-
building blocks themselves were typically also waveguide-based and erally limited by resistor–capacitor effects of the electrical circuitry,
would therefore take advantage of optical field confinement in compact which scale with the length. Moreover, for long devices, the modulator
cross-sections over distances many orders of magnitude beyond the cannot be seen as a lumped electrical element and a mismatch between
Rayleigh-range of optical beams diffracting in free space. the spatial distribution of the modulated optical field and the applied
For several decades most of the research efforts would focus on the electric field can occur as a result of the difference between the group
key building blocks: passive optical functions such as spectral filtering, velocities of the optical and electric wave respectively. In this context
electro-optical modulation, optical amplification, laser emission, light the highest bandwidth designs almost invariably use a traveling wave
detection, and more. Given the difficulty of achieving all these func- approach for the electric drive signal, such that it has the same propa-
tions in a single material system, the field diverged. For purely passive gation speed as the optical wave. With this approach modulation band-
functions low loss dielectric materials were chosen, with silica being an widths up to about 40 GHz have been reported.
important example. The term “planar lightwave circuit” (PLC) was typ- The status quo as described above has been broken recently through
ically used for such passive chips. For light amplification and laser the use of high-index-contrast photonics. Indeed, the waveguides de-
emission III-V semiconductors were the primary material of choice, scribed earlier were typically low-index-contrast waveguides in which
with the exact material choice dictated by the wavelength of operation. in-diffused ions provided the (very weak) index contrast between core
Here one would take advantage of the power of epitaxy to create layer and cladding, with a weak optical confinement as a result (Fig. 1). In
structures that would at the same time provide confinement for free Ref. 2 a drastically different approach is proposed and demonstrated.
charge carriers (electrons and holes) as well as for photons. For optical The LN-waveguide is a thin and narrow strip of LN embedded between
phase modulation one would resort to electro-optic materials with a silica claddings. This high-index contrast waveguide ensures a very
strong Pockels effect. The Pockels effect is an ultra-fast and optically compact guided mode cross-section, as a result of which the electrodes
broadband electro-optic effect in which an electric field induces can be positioned much closer to the waveguide. As a result, for the
changes to the relative position of bound electrons and atomic nuclei, same applied voltage a much stronger field can be induced without
which translates into a refractive index change that scales linearly with inducing a high optical loss. Additionally the design allows for inde-
the applied electric field. This allows for pure phase modulators. The pendently tuning the speed of the light and the electrical travelling wave
Pockels effect can only occur in materials that lack inversion symmetry, without sacrificing the overlap between the electric field and the optical
such as in non-centro-symmetric crystals (like LN) or poled polymers. field. The high mismatch between the dielectric constants at optical and
microwave frequencies is tackled in a traditional LN modulator by
*Address all correspondence to Roel Baets, E-mail: [email protected]
introducing a silica dielectric layer between the LN and the electrodes.
This however reduces the overlap. In Ref. 2 the velocity of the micro-
© The Authors. Published by SPIE and CLP under a Creative Commons Attribution
4.0 Unported License. Distribution or reproduction of this work in whole or in part
wave is adjusted by tuning the buried oxide layer’s thickness which has
requires full attribution of the original publication, including its DOI. [DOI: 10.1117/1 very little influence on the optical mode. The resulting device boosts an
.AP.1.3.030502] impressive performance with a V π Lπ -product between 2 and 3 V · cm,

Advanced Photonics 030502-1 May∕Jun 2019 • Vol. 1(3)

Downloaded From: https://www.spiedigitallibrary.org/journals/Advanced-Photonics on 27 Sep 2019
Terms of Use: https://www.spiedigitallibrary.org/terms-of-use
 COMMENTARY

(a) (b) (c)

Metal contact SiO2 LiNbO3 Ti:LiNbO3 Si BCB

Fig. 1 (a) A traditional lithium niobate modulator; (b) a cross-section of the modulator of Ref. 2;
(c) a cross-section of the modulator of Ref. 1.

a modulation bandwidth up to 100 GHz and a device loss well below of amorphous materials the co-integration can be done by straightfor-
0.5 dB. While the used approach may seem relatively straightforward, ward thin film deposition (sputtering, plasma-enhanced deposition,
it has taken non-trivial technological advances to make it happen. atomic layer deposition, etc.), but in the case of crystalline materials
The crystalline thin film of LN is manufacturing by means of wafer one needs to resort to more sophisticated methods. One may try to
bonding to form a so-called LN-on-insulator wafer (NANOLN). The use an epitaxial growth technique directly on silicon (or a deposition
etching of ridge waveguides is notably difficult in the case of LN technique followed by anneal to create a crystalline phase), but this is
and takes a carefully optimized reactive ion etching process in order typically very demanding and only possible for certain material com-
to yield low-loss waveguides. binations. Invariably it also involves a process step at elevated temper-
However impressive the performance of this result is, the LN-modu- atures. The alternative is to start from a crystalline thin film grown onto
lator is still a discrete component. In the past few decades the field of a different host substrate and then somehow transfer and bond that film
photonic integration has been moving in the direction of platforms that to the silicon photonics wafer. Methods to do so have become relatively
enable the integration of photonic circuits with a diversity of passive mature in recent years. Either they use bonding of the thin film to the
and active components within one circuit. This is a non-trivial task be- silicon wafer with subsequent removal of the host substrate4 or they use
cause different materials are good at different functionalities. Therefore lift-off of the thin film (and possibly even structured devices) from the
integrating multiple functionalities within one materials platform typ- host substrate followed by bonding to the silicon wafer. The latter tech-
ically leads to compromises on performance, at least in an early phase. nique is known as micro-transfer printing.5 These techniques have led
Nevertheless the benefits in terms of cost and reliability resulting from to many impressive demonstrations of co-integration of III-V devices—
the integration on one platform are important enough to make it highly in particular laser diodes and semiconductor optical amplifiers—with
relevant from an industrial point of view. Moreover, when many re- silicon PICs. But also a variety of other crystalline materials have been
search and industrial actors focus their efforts on one platform, a vir- co-integrated with silicon PICs.
tuous circle arises in which advancement of performance accelerates In Ref. 1 the co-integration of LN-modulators with silicon wave-
and economy-of-scale effects create an ecosystem in which fabs and guides is done with the help of die-to-wafer bonding. Although such
fabless users can thrive. an approach has been demonstrated before,6 the performance here is
While the notion of a one-platform-approach has been tried with boosted. Instead of relying on hybrid modes where part of the light
multiple materials, it is quite clear by now that silicon photonics is gain- is in the silicon and part of the light is confined in the lithium niobate
ing the largest momentum. Silicon photonics makes use of the tool sets layer, the authors make sure all of the light is in the lithium niobate
and process flows found in CMOS fabs to manufacture photonic inte- layer. They achieve this with vertical adiabatic couplers. These consist
grated circuits with silicon of silicon-nitride-based waveguides, inter- of a silicon inverted taper and a superimposed straight, LN waveguide.
connecting a diversity of components. The most common modality is to Due to the high confinement in the LN waveguides they are able to
use silicon-on-insulator (SOI) wafers in which passive silicon wave- make efficient LN modulators with a V π Lπ = 2–2.5 V · cm while
guides are formed along with waveguides with p-n junctions to form the routing was done in the silicon waveguide layer. Also in this study
phase modulators, based on the modulation of the free carrier concen- a travelling wave design was achieved. This allowed for an electro-optic
tration, as well as monolithic germanium detectors. With this combi- bandwidth of 70 GHz.
nation a variety of impressive products has been developed, in These results show the potential of how miniaturization enabled
particular high symbol rate (at least up to 40 Gbaud) optical transceiv- by high index contrast can enhance the performance of devices.
ers for the telecom and datacom market. By using the high confinement in thin film layers very efficient mod-
Silicon and germanium do lack the proper material characteristics ulators result. Moreover, by relying on a heterogeneous integration
however for all functionalities that one may wish for. Not only is light technique, Ref. 1 shows the integration of LN on a silicon photonics
emission highly inefficient, also the Pockels effect is absent. Therefore platform. These results pave the way to modulators that can operate
it is not possible to build a pure phase modulator in a similar way as is at very low voltages at high speeds on silicon. However, while LN is
the case in LN. While the free-carrier-based modulators mentioned maybe the best known electro-optic material, it is certainly not the only
earlier perform quite well, they have intrinsic limitations in speed and material that can be integrated on a silicon photonics platform. Other
optical loss. Moreover, they do not allow for pure phase modulation research teams have demonstrated high speed modulators that are based
because there is residual amplitude modulation due to free carrier in- on highly electro-optic polymers,7 layers of barium titanium oxide
duced loss. Already today they are used more or less at a performance (BTO)8 or lead zirconate titanate (PZT).9 These materials have even
level close to physical limits, but the market is calling for even higher a stronger electro-optic effect as compared to LN. The future will tell
performance so as to address the insatiable demands for bandwidth. which of these materials will end up in the new generation of high
This is where heterogeneous integration comes in, whereby a diversity speed transceivers as well as in other applications that require low loss
of materials is co-integrated on a silicon photonics platform. In the case phase modulation.

Advanced Photonics 030502-2 May∕Jun 2019 • Vol. 1(3)

Downloaded From: https://www.spiedigitallibrary.org/journals/Advanced-Photonics on 27 Sep 2019
Terms of Use: https://www.spiedigitallibrary.org/terms-of-use
 COMMENTARY

References with IMEC. He has made a broad variety of contributions to the field of
photonic integrated circuits, both in silicon and in III-V semiconductors.
1. M. He et al., “High-performance hybrid silicon and lithium niobate His current research interests focus on photonic integrated components
Mach–Zehnder modulators for 100 Gbit s–1 and beyond,” Nat. for new application areas, such as smart sensors and biomedical instru-
Photonics 13(5), 359–364 (2019). mentation. He has been granted several scientific prizes and is a fellow
2. C. Wang et al., “Integrated lithium niobate electro-optic modulators of IEEE, EOS, and OSA. He is the director of the Center for Nano- and
operating at CMOS-compatible voltages,” Nature 562(7725), Biophotonics at UGent. He has been awarded an ERC Advanced Grant
101–104 (2018). and has received a Methusalem Grant from the Flemish government.
3. S. E. Miller, “Integrated optics: an introduction,” Bell Syst. Tech. J. He coordinates ePIXfab, the European Alliance for Silicon Photonics.
48(7), 2059–2069 (1969).
4. S. Keyvaninia et al., “Ultra-thin DVS-BCB adhesive bonding of Bart Kuyken received the BS and MS degrees in electrical engineering
III-V wafers, dies and multiple dies to a patterned silicon-on- and the BS degree in applied physics from Ghent University in 2008. He
insulator substrate,” Opt. Mater. Express 3(1), 35–46 (2013). received the MS degree in electrical engineering from Stanford University
5. A. De Groote et al., “Transfer-printing-based integration of single- in 2009. In 2013, he received the PhD degree from Ghent University for
mode waveguide-coupled III-V-on-silicon broadband light emitters,” the thesis “Four-wave-mixing in dispersion-engineered silicon nanopho-
Opt. Express 24(13), 13754–13762 (2016). tonic circuits for telecommunication and sensing applications.” As a post-
6. A. J. Mercante et al., “Thin film lithium niobate electro-optic modu- doc, he worked in the field of nonlinear optics. He was a visiting scientist
lator with terahertz operating bandwidth,” Opt. Express 26(11), at the IBM TJ Watson Research Center (U.S.) in 2011 and 2012,
14810–14816 (2018). and at the Max Planck Institute for Quantum Optics (D) in 2013 and
7. L. Alloatti et al., “100 GHz silicon-organic hybrid modulator,” 2015. His work includes the integration of nonlinear optical functions in
Light Sci. Appl. 3(5), e173 (2014). silicon photonics waveguides, 2-D materials, and THz waveguide integra-
8. S. Abel et al., “Large Pockels effect in micro-and nanostructured ba- tion. In 2015, he was appointed an assistant professor in the Faculty of
rium titanate integrated on silicon,” Nat. Mater. 18(1), 42–47 (2019). Engineering of Ghent University. He has been granted an ERC Starting
9. K. Alexander et al., “Nanophotonic Pockels modulators on a silicon Grant for the development of on-chip frequency combs.
nitride platform,” Nat. Commun. 9(1), 3444 (2018).

Roel Baets is a full professor at Ghent University and leads the Photonics
Research Group of Ghent University, a research group associated

Advanced Photonics 030502-3 May∕Jun 2019 • Vol. 1(3)

Downloaded From: https://www.spiedigitallibrary.org/journals/Advanced-Photonics on 27 Sep 2019
Terms of Use: https://www.spiedigitallibrary.org/terms-of-use