Emerging Technologies 2018 Session Listing
The program is subject to change in the weeks leading up to the conference. Check back here for the latest schedule, or follow us on Twitter for real-time notice of updates to the program.
Session C2: Sensors
Start Time: 09:00, Thursday, May 10
Room: Diamond Head
Chaired by Fabio Di Francesco, Università di Pisa (firstname.lastname@example.org)
- 9:00 How-Foo Chen, National Yang Ming University
with C-Ha. Chen and P-B. Wang
Designing and fabricating a medical surface plasmon resonance biosensor: application on antimicrobial susceptibility test of E. Coli
- 9:20 Sigurd Wagner, Princeton University
with T. Moy, Y. Afsar, L. Aygun, Y. Mehlman, J.C. Sturm and N. Verma
Thin-film circuits for interfacing large-area sensor arrays and CMOS circuits
We have been demonstrating sensor systems that combine large-area arrays of sensors, made in thin-film technology, with CMOS ICs . We foresee that such systems will become unobtrusive components of the built environment, with the purpose of augmenting human sensing. Systems for sensing mechanical strain , gestures , images , sound , and electrophysiology (EEG)  have been demonstrated. Our goals are (i) to understand the optimal distribution of functions between the large-area thin-film and CMOS domains, and (ii) to explore the application space for large-area sensor systems made with this hybrid technology. A priority has been to reduce the number of electrical interfaces between the thin-film and the CMOS domains. Interfaces have been reduced by using purely circuit-based approaches, and also by introducing algorithmic techniques. We will describe several of the thin-film circuits developed for this purpose.
 N. Verma et al., Proc. IEEE 103, 690 (2015).  B. Glisic et al., Proc. IEEE 104, 1513 (2016).  Y. Hu et al., IEEE Custom Integrated Circuits Conf., 2014.  W. Rieutort-Louis et al., IEEE J. Solid-State Circuits 51, 281 (2016).  J. Sanz-Robinson et al., IEEE JSSC 51, 979 (2016).  T. Moy et al., IEEE JSSC 52, 309 (2017).
- 9:40 Lado Filipovic, Technische Universität Wien
CMOS-compatible semiconductor-based gas sensors
Recently, there has been an ever-increasing demand for functional integration in a single device. Connecting multiple technologies using bond wires can negatively impact performance due to the associated increase in circuit resistances. The highest efficiency is reached when all functionalities are fabricated on a single substrate, deemed System-on-Chip. Fabrication on silicon allows for the efficient integration of sensors and CMOS structures into a truly monolithic device.
The integration of a metal-oxide (MOX) based gas sensing device into silicon technology is a particular challenge. The sensing layer, which can be a nanowire, nanosheet, or a thin film, must be heated to temperatures between 300°C and 500°C to operate as a sensor. For this reason, a microheater is implemented underneath the sensing element. The heater is the main power dissipater and its design determines the total power consumption of the device.
Our work is focused on understanding the consequences of the complex microheater structure on the mechanical stability of the sensor and the optimized operation of the MOX layer. We will describe how the reliability and gas-MOX interaction can be modeled and optimized for low-power operation. Furthermore, we will present some 2D semiconductor alternatives, which go beyond the limitations of MOX-based sensors.
- 10:00 Yves-Alain Peter, École Polytechnique de Montréal
Gas sensing with optical microresonators
We present an optical nose on chip made of a matrix of optical gas sensors. Optical noses integrated on chip present numerous advantages over electronic noses such as low power requirements, robustness, and immunity to electromagnetic fields, remote sensing and lower price. Miniaturized on-chip sensor, designed to detect air-borne compounds, are essential for inexpensive monitoring systems that are portable and deployable on a large scale. The optical sensing device is based on a reversible absorption of a gas with a dedicated polymer matrix. The sensor is integrated on chip and is therefore small, compact, and can be distributed in a network enabling low power consumption. We demonstrated that it can monitor several volatile organic compounds (VOCs), such as alcohols, aldehydes, ketones, carboxylic acids, amines and aromatics, that it operates in a reversible fashion, under different environmental conditions (temperature and humidity), and that it detects concentrations in the order of parts per million (ppm).
- 10:20 COFFEE BREAK
(Mt. Curie Foyer, Sutcliffe Foyer)
- 10:40 Chi Xiong, IBM
Monolithically integrated silicon photonic gas sensors
- 11:00 Justin Caram, University of California, Los Angeles
Probing new chemistry in the shortwave infrared using superconducting nanowire single photon detectors
The short-wave infrared (SWIR or NIR II), is the sparse spectral region from 1-2 microns (0.6-1.2 eV), red of most electronic transitions, yet blue of many infrared active molecular vibrations (except C-H and O-H overtone combination bands). While the primary use of the SWIR has been in fiber-optic communications, recent research has realized its potential in deep tissue imaging, biometric identification, satellite telemetry for weather and plant cover, and pedestrian imaging for self-driving cars. However, there has been limited work combining SWIR with high temporal and spatial resolution methods such as confocal fluorescence imaging and TCSPC, commonly used in chemistry and biological questions. My group uses superconducting nanowire single photon detectors (SNSPDs), to extend a suite of powerful photon correlation methods from the visible into the SWIR. This research enables unique measurements of weakly emissive states critical to the development of optoelectronic materials and devices. We explore the photophysics of a broad range device-relevant inorganic and organic semiconductors, whose dynamics depend strongly on the properties of states that emit in the SWIR, including triplets, charge-transfer states and defects.
- 11:20 Seiji Kajihara, Kyushu Institute of Technology
A full digital temperature and voltage sensor for field testing
In this talk we present a novel digital sensor to measure temperature and voltage on a VLSI chip simultaneously. The sensor is composed of ring oscillator based logic circuits, and all computations are done with fully digital process. Compared with the conventional analog sensors, the proposed sensor has several technical merits which are small chip area, quick response time, low power dissipation and no need to prepare A/D convertor and reference current/voltage. Therefore, more than one sensor can be placed at various locations on a chip. In addition, the sensor can have an aging-tolerant structure for electro-migration, BTI and HCI. In order to reduce the influence of process variations on measurement accuracy, a calibration method that uses an initial measurement value of each sensor is adopted. In order to estimate measurement accuracy of the sensor, experimental results using circuit simulation and a fabricated test chip are also presented where we investigates effectiveness of the sensor derived from reduction of temporal and spatial variations. The comprehensive evaluations show that the total measurement error is smaller than the analog sensors, and it implies the importance of real time and contiguous measurement.
- 11:40 Bhaskar Choubey, University of Oxford
Increasing the M/NEMS Sensors population per chip
Micro/nano sensors designers have generally avoided placing more than a few sensing devices on the same chip fearing coupling and other undesired effects. The aim has typically been to improve the sensitivity, functionality, manufacturability and often size of single or few devices per chip or die. Often, the contacts are significantly larger than the actual device being built, wasting precious microfabrication area. On the other hand, we have always preached that our devices will end up costing the same as silicon VLSI devices, as we make them in similar foundries. However, having just one or few devices per chip reduces the potential functionality and does not align with VLSI philosophy. More importantly, most of our devices follow the philosophy of "one design, one fab and one product", which is not in line with VLSI industry wherein one foundry caters to a large number of designs. This means that the cost per device of a MEMS sensor is still very high. In VLSI domain, more than one device are placed on a single chip, by providing suitable interconnects and thereby limiting the needs of electrical external contact pads. Providing several M/NEMS devices on a chip leads to generally undesirable collective behaviour. However, if we can utilize this undesirable behaviour, we can potentially provide for several M/NEMS sensors on the same chip. In this presentation, we will present mathematical techniques, physical realisations and applications of utilizing such collective behavior leading to large number of M/NEMS sensors on a chip, yet with limited contact counts or increase in bandwidth.
Support from our sponsors allows us to provide additional special events and networking opportunities during our conferences.