Research: An Ultra-compact Nanophotonic Optical Modulator using MultiState Topological Optimization
UniversityWafer, Inc. supplied the 100mm Silicon waferso to fabricate modulators.
We fabricated the NOM device on the conventional platform of a silicon-on-insulator (SOI) wafer. As per our design constraint, the thickness of the top silicon layer was required to be 300 nm. However, since we were unable to procure SOI wafers with this exact silicon layer thickness at the time, we resorted to preparing our own wafer. In order to do this, first a 4-inch silicon wafer (University Wafers)was cleaned using acetone, methanol and isopropanol.
Graphene-Based Silicon Modulators
Graphene-based silicon modulators are among the latest developments in the field of photonics. They have been developed by using a doping compensation method to reduce the number of free carriers. Their work has resulted in a high phase efficiency of 3.5 V.cm and 40 Gbit/s modulation rate. They are the perfect candidates for next-generation photonics. But how do they work? And are they really better than existing modulators?
The first cross-section of a Mod1 silicon modulator is shown in Figure 1A. In this structure, p-type silicon and n-type silicon are highly doped to form ohmic contacts between their coplanar waveguide electrodes. The second cross-section of a Mod1 silicon modulator has a p-n junction positioned to the right of the rib waveguide to maximize carrier depletion movement.
The University of Sydney and SEAS researchers fabricated a prototype of a Mod1 silicon modulator with a 3.5-mm MZM and a 4V RF signal. They observed that the signal was stable even at high optical intensities. This technology could potentially support modern communications in data centers and satellites, as well as a future quantum internet. The researchers worked with colleagues from Harvard University's Center for Nanoscale Systems and the University of Sydney's Quantum Information Technology Center.
The Mod1 silicon modulator is based on the Free-Carrier Plasma Dispersion (FCPD) effect, a mechanism by which carriers change the refractive index of a semiconductor. This process also alters the absorption coefficient and propagation velocity of light. It is this effect that allows SOI-based modulators to be manufactured. The Mod1 silicon modulator may advance the field of quantum communications while facilitating the integration of traditional electronics.
A P-type layer 112 is provided on a micro-resonator 102, while an N-type layer 118 is formed on the other side of the modulator section. The PN junction between P-type layer 112 and N-type layer 118 may be vertical or substantially normal to the central axis 126. In addition to P and N-type layers, a material 110 is used between the P and N-type layers.
The photonic modulator 100 may include one or more modulator sections 108-1. Each of these sections may subtend a circumference of the micro-resonator 102 and are spaced approximately p radians apart. The voltage applied to these modulator sections will modulate the light. The polarity of the applied voltage will control the amount of carriers that are added or removed from the micro-resonator 102.
Optical simulations of the Mod1 silicon modulator have shown that the simulated efficiency closely matched the measurements. These experiments demonstrate the potential of this silicon modulator for use in communication devices. The Mod1 silicon modulator is an effective tool for achieving high-speed data transmission. In fact, it is one of the most important advances in semiconductor technology. So, the next time you need to make a decision about the efficiency of the Mod1, take advantage of this technology.
During measurements, the center bias point on the modulator sections 108 in FIG. 1A may be adjusted to make the diode run on the edge of Vto and Vbd. Note that this effect is largest when the voltage is close to Vto. The depletion capacitance C is proportional to (Vbi-Va), so it may affect the diode's operating conditions. This technique is highly advantageous for applications such as digital audio transmission.
A simulation model based on the '1' state is designed to demonstrate the effect of imbalance arms on the output voltage. The measurement was done using a 4V reverse DC bias and a simulated RF signal. A random 40Gbps bit sequence was used to generate the RF signal. Both simulations and measured results show that the Mod1 silicon modulator has a high phase shift. In addition, the Mod1 silicon modulator is compatible with VML and CML signaling systems.
A Mod1 silicon modulator has undergone a remarkable transformation in the last four years. The device size can be reduced to a micrometer and operates at market-standard rates of up to 40Gb/s. It is now possible to integrate silicon optical circuits with electronic circuits. Its high-speed operation allows chip-to-chip communications and networking. Its lack of electric field-based modulation mechanism makes it difficult to replace the conventional optical modulator.
A fabricated Mod1 silicon modulator may include a P-type contact region 116 and an N-type contact region 118. These regions may be coupled to differential drive circuitry 200. The circuitry may have one differential output. This differential drive circuitry can be connected to one of the P-type layers 116 and 118 for controlling modulator sections. Then, the P-type and N-type contact regions of the Mod1 silicon modulator are driven by a differential driving circuitry 202.
Graphene-Based Silicon Modulators
Graphene-based silicon modulators were first demonstrated at a conference in 2015. They were used to improve the efficiency of silicon-based ICs and to achieve a better balance between power and area. They have demonstrated a modulation depth of about 38% using a single layer of graphene. A double-layer graphene modulator had a modulation depth of about 13 dB.
Graphene modulators work by splitting the graphene segment into two equal segments. The segments determine the frequency response of the modulator. The ratio of the two segments defines the extinction ratio of the modulator. The graphene segments can also be segmented using binary or thermometer coding. The frequency response is further defined by the number of segments. Further, graphene segments are programmable, allowing the device to be programmed with a desired frequency.
The proposed device has ridge wave guides with a width of 500 nm and a height of 220 nm. The device also features an interaction region of about 40 mm. The optical images of the device can be found in Fig. 3 (a). The device includes electrode pads and an interaction region. A planform SEM image of the interaction region reveals that the graphene layer is above the silicon waveguide and over the electrical pads.
Graphene-based silicon modulators can have a very high modulation depth. This allows the device to have a small footprint while still achieving high modulation depth. High modulation depth is achieved by increasing the peak-to-peak gate voltage swing. A graphene-on-silicon structure enables a three-dB modulation depth. There are also many challenges with this design, however.
Because graphene has low parasitic pad capacitance, it is possible to achieve a large-scale inductance and still maintain high performance. However, this combination is not yet mature enough to be used for commercial modulators. Graphene-based silicon modulators should be available in the market before 2020. A good way to start is to develop a prototype model for this device. You can find more information about graphene-based silicon modulators on the official website.
To design a functional hBN-encapsulated graphene-based EA modulator, we fabricated graphene-encapsulated hBN electrodes. Graphene electrodes are protected by HfO2 and hBN. In the simulations, we determined the Fermi energy of graphene and measured the transmission curves. Graphene modulators can achieve high modulation rates at low drive voltages.
Graphene-based silicon waveguide modulators have demonstrated high linearity in microwave photonics. However, simulation results and real-world measurements differed. This may be due to fabrication errors or to a degradation of the graphene layer after the transfer. However, these modulators are promising for various integrated microwave photonic applications. With the right configuration and proper measurements, these devices could be a good match for current and future microwave photonics equipment.
The high Q of the designed modulator was derived from the intrinsic properties of graphene. They show high modulation depth and low insertion loss. This modulator can operate at a drive voltage of about 1 V pp and at a bit rate of about 136 GHz. However, the short duration of the modulation intervals may limit the practical application of such devices. Therefore, they are ideal for use in communications.
The experimental results of graphene-based silicon modulators demonstrate that they are feasible. They also show good agreement with theoretical predictions. The parameter Re(neff) increases as the chemical potential increases from 0 eV to 1 eV, and then decreases again. This indicates that a two-stage wavelength shift can occur in graphene. The mode switch occurs around 1 V. Thus, a graphene-based silicon modulator can be used for optical communications.
The graphene-based silicon modulator has a high maximum Fermi energy, with the dielectric permittivity and strength being the key determinants. Graphene-based silicon modulators exhibit a modulation efficiency of around 2.2 dBV-1, compared to other modulators. A typical graphene dielectric is made of HfO2 or hBN. These dielectrics exhibit inhomogeneous doping, but do not have a high electronic mobility.
Graphene-based silicon modulators have been developed to achieve efficient phase modulation. These devices can be used in Mach-Zehnder interferometers, microring resonators, and add-drop microring resonators. Graphene-based silicon modulators are compatible with CMOS technology. They have been demonstrated for use in nonlinear signal processing and in sensors.