Community Research and Development Information Service - CORDIS

ERC

AAATSI Report Summary

Project ID: 278794
Funded under: FP7-IDEAS-ERC
Country: Netherlands

Final Report Summary - AAATSI (Advanced Antenna Architecture for THZ Sensing Instruments)

Integrated Technology revolution
I started University in 1986. The length of the CMOS (Complementary Metal Oxide Semiconductor) gates at that time was 1 micro-meter and it is now 10 nano-meters. For the largest number of applications and engineers, Moore’s law meant increasingly smaller computers and now tablets. However, for microwave engineers it means low cost sources and receivers waiting to be exploited. When I started following the integrated technology trend, the maximum frequency of the CMOS transistors was 1 GHz (Fmax=1GHz). By the time I finished the PhD, 2000, we were at 200 GHz. We now have CMOS technology that provides silicon transistor with Fmax=1 THZ. A factor 1000 in maximum frequency of operation in 30 years. This fact has nothing to do with my contribution. But it explains why it is was so timely to invest in THz research. Moore’s law has finally opened a new era in the field of electronic: THz spectrum is de facto available for widely spread applications.

Wide Band vs. Dispersion
The main potential built of the THz field for electronic applications is the availability of an enormous frequency spectrum. Despite this potential and the present availability of low cost electronic sources, the THz domain still presented (6 years ago) important challenges. Specifically, the propagation losses in metals at THz frequencies are dramatic when compared to low frequencies. For this reason, and because of the limited information capacity of the present (and future) analogue to digital converters, the wave manipulation that at low frequency is performed in the electronic circuitry, in the THz domain can only be realized analogically by means of quasi-optical components. Thus any efficient sensing or communication schemes in the THz domain will have to be based on immediate conversion of the very wide-band THz signals from guided to radiated and vice versa: antennas integrated with active components and detectors.
Before this grant the major problem that the THz field was facing was, in my view, the arising of dramatic losses that all known antennas introduced in the THz budget links for wide band signals. These losses were either built in the form of poor beams quality or in the form of unacceptable distortion of the signals that could be recovered in the electronic front ends. In one word, the main issue was dispersion, i.e. the mechanism by which an electromagnetic signal, short in time, thus broad in frequency domain, is damaged as it propagates through a high-frequency electronic front end or channel. Dispersion was the bottleneck that THz electronic was facing and that this grant was targeted to solve.

Goal of the Project: MTIA (Monolithic Terahertz Integrated Antennas)
Accordingly the goal of the project was to revolutionize THz sensing systems by introducing breakthrough antenna technology that would solve the dispersion bottleneck. The reason for focusing on sensing, rather than communication was simply that sensing schemes that involve extremely large bandwidth were already being proposed in space science and security scenarios. Thus the technological breakthroughs could be easily quantified and evaluated by introducing them in target radiometric systems, whose performances are directly proportional to ‘Bandwidth × Efficiency’. These systems are Kinetic Inductance Detector (KID) arrays for space-based spectroscopy, GaAs Time Domain Sensing systems, and CMOS integrated focal plane imaging cameras.
At the time of the proposal, I had already identified two basic theoretical breakthroughs (leaky lens antennas and connected arrays) to be introduced in the antennas used in the target THz sensing systems. I identified the third one (artificial dielectric layers) in the period that ran from the proposal submission to the grant award.
Viewed in a coherent picture these separate antenna breakthrough are certainly helping to shape a natural transition in integrated technology for electronics. Possibly the natural acronym for high frequency circuitry will evolve from the present MMIC (Monolithic Microwave Integrated Circuits) to MTIA (Monolithic Terahertz Integrated Antennas) to highlight that as frequencies become higher, it will all be about antennas rather than transmission lines.

Creation of the THz Sensing Group
Right after the grant was awarded, the THz Sensing Group was established, as part of the Microelectronics department, in the faculty of Electrical Engineering, Mathematics and Computer Science at TU Delft. It now includes 30 scientific members among which 7 professors (two part time) , 10 PhD students, 2 postdocs and and 5 permanent visitors (employed by companies and research centers) and 5 master students. I was the only person affiliated with TU Delft at the start of the group. While the ERC grant mainly covered the salaries of 2 PhD students and 2 postdocs, it is fair to say that a lot of its research activities rotated for a long time around the AAATSI grant and benefited from its coordinated effort. The group now hosts otrher two active ERC grants, and it is possibly the world leading group group in THz Sensing. An up to date description of the THz Sensing Group is available online at the link (http://terahertz.tudelft.nl).

Results of the project
At the end of the project virtually all the anticipated goals of the grant have been achieved even if, as anticipated, using many more resources than those associated to the financial means of the grant. These resources emerged by rendering the strategy of the grant the strategy of the entire Terahertz Sensing Group.
The project has seen two virtually parallel completely successful tasks: on one hand, we have completed the development of a theoretical framework to represent efficiently the electromagnetic field within focusing systems composed of thousands of elements that need to be operated over extremely large bandwidth; on the other hand, we have developed the prototypes of the radiator arrays that demonstrate that the promised breakthrough concepts were actually feasible at THz frequencies. The concepts of leaky lens, connected array and artificial dielectrics enhanced MTIA have all been demonstrated at THz frequencies.
Being able to actually manufacture the building blocks of the antenna architectures was anticipated to be the most risky part of the entire project. In fact, as applied as we try to be, at the beginning the THz Sensing group was only a group of electromagnetic theorists. We had to build solid alliances with other groups (some in our Micro-electrionics Department (http://microelectronics.tudelft.nl/), some in the Netherlands (SRON-Netherlands Institute for Space Research) and some in the United States) in order to develop the desired micro-fabrication capabilities.
On the leaky lens side, especially thanks to the alliance with SRON we have reached the manufacturing, demonstration and characterization of many arrays of thousands of dual polarized, wide band antennas with efficiencies higher than ever reported for even a single antennas. The sensitivities allowed by these arrays effectively shift the target of Deep Space Investigation front ends from 10% relative bandwidth (BW) to decade BW’s, rendering THz on chip wide band spectrometry a feasible field.
On the Artficial Dielectric side we have developed a clean room process that is able to double the efficiency of antennas integrated on silicon chips bringing it to essentially 80%, virtually independently from the BW of operation. This procedure has been demosnstrated at 300 GHz but also for much more useful 80 GHz front ends, has been patented and exploited by one of the biggest semi-conductor industry players in Europe (NXP). These efficiency levels mean CMOS THz front ends are effectively usable and not only on paper.
Finally on the connected arrays side, there has been the realization of pulsed power array sources, producing a 1mw of power spread between 0.1 and 0.8 THz. These power levels are at least 1 order of magnitude higher than the power available from commercial providers, or even reported as reproducible results. These power levels open the way for THz radars at essentially no cost with respect to the state of the art.

Reported by

TECHNISCHE UNIVERSITEIT DELFT
Netherlands
Follow us on: RSS Facebook Twitter YouTube Managed by the EU Publications Office Top