Skip to main content


Periodic Reporting for period 2 - Nanonets2Sense (Nanonets2Sense)

Reporting period: 2017-08-01 to 2019-03-31

Nanonets2Sense partners joined their force to overcome one of the main roadblocks towards the widespread use of nanowires (NWs) in sensing applications: their integration cost. It explored a new approach, where random networks of NWs, called nanonets (NN), were used as sensing material. The starting breakthrough originated from our capability to obtain conducting NNs by sintering at less than 400°C. On this basis, we developed an innovative concept for 3D sensor integration.
The consortium included well recognized European partners, including academic laboratories, an SME and a large company, representing the whole chain from basic and applied research to foundry and products development, ensuring that the approach would combine sounded physical concepts with industrial vision. The project benefited from their strong and complementary expertise to address all material, device, and circuit aspects. As typical applications, it targeted the detection of DNA strands in body fluids and of acetone in breath, as possible biomarkers in oncology or for sugar metabolism, respectively.
Thanks to intense coordinated efforts, Nanonets2Sense achieved its core objectives. It was able to establish a technology for the 3D above-IC integration of NN-based sensing devices on a CMOS platform, and delivered two demonstrators.

Nanonets were obtained low cost nanowire growth and assembling. The controlled growth of both Si and ZnO nanowires was achieved, with independent control of diameter and length.

For DNA detection, we developed a low temperature process (<400°C), compatible with above CMOS integration, for the fabrication of back-gated NN-based field-effect transistors (NNFETs). Functionality was demonstrated, with several decades of current swing. For given density, an optimum was found in terms of device geometry. The active material was made of a silicon NN, functionalized with DNA probes. Reference NWFETs, fabricated with a top-down technology, were also provided by KTH and tested as sensors for comparison. A new eco-friendly surface functionalization process was successfully developed, which respected metal contacts integrity. The detection scheme was based on the measurement of the threshold voltage shift resulting from target/probe hybridization. Ultra-low power circuit design techniques were applied, in order to guarantee low power consumption, which is very important in portable point-of-care devices. The circuit was fabricated with ams AG 0.18µm technology and successfully tested. Two mask sets were designed to enable joint processing, the readout being processed by AMS, the pre-patterned back-gate and contacts being integrated by KTH and the final NNFET processing being handled by GINP. The integrated sensor allows sequential readout of a matrix of 32 identical sensors organized in 4 groups of 8, so that each group can be functionalized with a different probe and perform some averaging.

For acetone detection, we used the ability of some metal oxides (MOx) to have their resistance changed in the contact of reducing of oxidizing gases. This process is thermally activated so that chemiresistive materials must be heated to respond. Controlled heating was obtained by a micro hotplate integrated in the locally thinned membrane substrate. The aim of Nanonets2Sense was to replace the standard sensing material in CCS devices with ZnO NNs. Thin membranes were found to be robust against processing. We developed an industrially-relevant processing procedure, where ZnO NNs were transferred and processed on coupons integrating micro hotplates. After dicing and packaging, these NN-based chemiresistors were characterized under gas. ZnO NNs demonstrated to respond to acetone. One of the keys has been the proper choice of a passivation layer, to improve material stability in time without compromising sensitivity to acetone. Finally, comparison in fair conditions suggests that microparticle and nanonet-based devices may be used together to extend the dynamic range.
Technological developments were supported by intense characterization and modelling activities. This was crucial, in order to understand the role of the networked structure of nanonets and to provide sounded TCAD models for circuit design. An original compact model was developed for Si NN-FETs, with the associated parameter extraction method, and was implemented in the circuit simulation platform. The dependence of NNFET parameters with channel length, nanonet density and temperature were discussed in terms of percolation effects, energy barrier at NW/NW junctions, and threshold voltage dispersion.

The project has also been active in identifying the many paths for future exploitation, under the leadership of its industrial partners and with the advice of a board of external physicians, scientists and device developers working in the field of medicine and health monitoring. Exploitation is strongly favoured by the fact that the proposed technology is generic. It has already started with a FlagERA project Convergence, which includes Si NN-based FET sensors functionalized with aptamers for the detection of thrombin. The models and the characterization and simulation methodologies developed in the project are also suitable for other types of percolating films.

Project results were disseminated through several channels. Information is provided at different levels on the project website. The scientific and technical outcomes of the project have already resulted in 21 peer-reviewed publications in international scientific journals or conference proceedings, and in 34 presentations at international conferences and workshops. Moreover, two demonstrators were built to showcase gas sensing in international forums. Dissemination to wider public was not forgotten, with 5 contributions in Austria, Sweden and France.
By using nanonets as sensing material, our synergetic approach retains the advantages of NW properties without the associated technological burden. With a smart combination of bottom-up and top-down technologies and a low processing temperature compatible with CMOS, it allows 3D integration into a compact sensor, where the sensing element, which is exposed to breath or biofluids, is integrated above the CMOS detection circuit, which is naturally protected. With an estimated power consumption of only a few dozen µW, the fully integrated DNA sensor has the potential to become a disruptive technology. The project also demonstrated that NNs can be integrated on micro hotplates with an industrially relevant process and that the resulting devices are functional as gas sensors whilst operating at temperature up to 400°C. It also highlighted that the engineering of NN passivation affords a powerful avenue to tune gas sensitivity and increase sensor lifetime.
Nanonets2Sense is thus providing a completely new technological building block to enhance CMOS chips functionality with biosensing capability. The impact is enhanced by the fact that the approach is generic and can be adapted to a large variety of NWs and target molecules. Solely in the medical field, where the availability of biosensors at low cost is a key for the widespread diffusion of point-of-care devices, applications can be envisioned in the fields of therapy monitoring, pharmacogenetics or metabolic disorders monitoring.