In the INTEGRATECH project we aim at demonstrating a new THz technology platform, based on the integration of graphene with novel electronic SiGe HBT/BiCMOS chip technology. Graphene possesses extremely high nonlinear coefficients, that allows one to upconvert sub-THz driving electrical signal into the THz frequency range. In this project we will combine this extreme THz nonlinearity of graphene with the novel BiCMOS technology, in order to efficiently generate THz frequency signals in an integrated graphene-on-chip device. In particular, we will use the chip technology to create a sub-THz driving field, and exploit the extreme nonlinear nature of graphene, in order to demonstrate on-chip THz frequency conversion. We believe that this novel THz technology platform will provide new avenues towards on-chip functionalities, such as THz generation, detection, and manipulation. It was recently discovered that graphene possesses gigantic nonlinear coefficients in the THz frequency range, largely surpassing that of any other known functional material. In particular, graphene allows for up-conversion of sub-THz input electronic signals to the THz range, with high harmonics generation efficiency reaching 1% per atomic layer. Further, graphene is compatible with existing, highly evolved semiconductor ultra-high frequency technology. Remarkably, the sub-THz driving field of the order of 10s kV/cm, needed for efficient generation of THz harmonics in graphene, is even one order of magnitude smaller than the channel field of a typical high-speed, sub-THz CMOS transistor. This suggests a clear route to novel integrated graphene-on-chip THz technology, combining CMOS or BiCMOS integrated circuits as a supplier of sub-THz driver signal, and graphene as an active up-conversion medium for the THz signal generation. The physical picture behind the extreme THz nonlinearity of graphene has been developed within the applying consortium, and is known as thermodynamic model of ultrafast charge transport in graphene. Further, the applying consortium has very recently demonstrated effective control of this THz nonlinearity in graphene via electrical gating. Finally, members of the applying consortium are also among the world leaders in novel SiGe HBT/BiCMOS ultra-high frequency integrated circuit technology. Therefore, the consortium behind this project possesses all the expertise necessary for its success.

THz imaging is an emerging field with main applications in non-destructive material analysis, biomedical applications, or security control. In the past 15 years, THz-cameras, capable of real-time imaging, have been developed using established Silicon CMOS or Silicon-Germanium BiCMOS technology. State-of-the-art Silicon CMOS THz cameras have 1024 pixels (32 x 32) on up to 10 x 10 mm2 large chips. Larger arrays with higher resolution are hardly accessible using conventional silicon technology because of high production costs, which scale directly with the die area. Furthermore, conventional rigid silicon technology does not allow constructing flexible focal plane arrays on curved surfaces reducing spherical aberrations, enlarging the field-of-view, and leading to a better off-axis resolution. We will develop flexible large area THz detector active matrix arrays using a backplane based on metal-oxide thin-film-transistor technology within the LATINO project. For THz detection,antenna coupled metal-insulator-graphene diodes will be used and optimized towards the specific requirements of the readout circuitry. Dedicated low-noise amplifiers and multiplexer circuits will be designed and fabricated using metal-oxide semiconductors. The goal is to demonstrate an active pixel based on thin-film-technology comprising an antenna, a diode, and an amplifier at the end of the first funding phase. In the second phase, the technology will be transferred to flexible substrates. Additionally, an array with 64 x 64 pixels will be realized on a 4 x 4 cm2 area. This will lay the foundation for a new class of high-resolution THz cameras going significantly beyond the limitations of conventional Silicon CMOS technology.

LATINO Team:

Prof. Thomas Riedl, t.riedl@uni-wuppertal.de
Prof. Daniel Neumaier, d.neumaier@uni-wuppertal.de
Prof. Ullrich Pfeiffer, ullrich.pfeiffer@uni-wuppertal.de

In the project MINTS it is planned to investigate and demonstrate electronic-photonic THz frequency synthesizer architectures which conform to the requirements of integration in silicon photonics (SiPh) and/or Indium Phosphite (InP) photonic technology. Building on the excellent spectral properties of mode-locked lasers (MLL) it is the target to achieve a jitter rsp. phase noise performance of at least one order of magnitude better than any purely electronic THz frequency synthesizer while achieving a similar or better frequency range and frequency resolution. The improvement in phase noise performance of the electronic-photonic THz synthesizers is achieved by locking electronic oscillators rsp. laser diodes onto the MLL signal, depending on the synthesizer architecture. We will investigate different MLL-based THz frequency synthesizer architectures both experimentally and analytically. Novel tuning techniques are proposed to achieve continuous frequency tuning. Effects of an integrated realization in InP and SiPh technologies will be studied analytically and experimentally. Our research will focus on synthesizer architecture studies and implementations in InP and silicon Photonic Integrated Circuit (PIC) technology. Since integration of the MLL itself is not in the focus of the proposed project, we will make use of commercial ultra-low-jitter MLLs from Menhir Photonics and Menlo Systems for synthesizer experiments. Together with ongoing research of other groups on integrated MLLs the project MINTS will pave the way for compact integrated THz frequency synthesizers in InP and SiPh technology with exceptionally low phase noise.