Final Report Summary - E-GNOSIS (e-Gnosis: a novel platform technology for quantitative mobile diagnostics) e-Gnosis: better mobile immunoassayse-Gnosis is an interdisciplinary project that combines state-of-the-art nanotechnology, electrochemistry and bioanalytical technology to develop a disposable biosensor chip that combines the low-cost and ease-of-use of lateral flow immunoassays with the sensitivity and multiplexability of quantitative lab-based systems. The ultimate goal is to obtain quantitative analyte measurements by placing a small sample droplet on the surface of the chip and then using a simple, reusable circuit for read-out (similar to the handheld potentiostats currently used for blood glucose measurements by diabetics). In the future, these features, along with a lower sample volume requirement than lateral flow assays, could allow the e-Gnosis platform to be used for more widespread screening and diagnosis in both professional and consumer-level settings such as medical diagnostics, health monitoring and agriculture.The two main project objectives are:1. Fabrication of multipixel chips 2. Chip-based immunoassays: Determination of chip performance characteristics by two novel immunoassay formats: a. Signal amplification using electroactive polymer nanobead labelled secondary antibodies b. Non-amplifying quantitation, using the size of the binding analyte to reduce peak current flow between the electrodes. Significant progress has been made towards the goal of chip-based electrochemical immunoassays. In the course of the MC project, working prototypes of the chip have been fabricated, packaged to allow connection to readout equipment and electrochemically tested. In addition, we successfully built, programmed and tested a low-cost potentiostat using a commercially available gas sensor chip with our electrochemical chips. Usually chips are encased in silicon nitride and thermoplastics to isolate them from humid or wet environments, as these can lead to failure by delamination. The e-Gnosis design depends on liquid access to the nanowell surface and from our previous work we had identified delamination as one of the main challenges in this project. We significantly increased the stability and lifetime of the chips in liquids, and have not seen any delamination below the bubble overpotential of the electrolyte with the new chip design. The chips are now stable in liquids for hours and survive repeated immersion, cleaning and drying cycles. We have tested a number of chip-packaging approaches, including flip-chip bonding, wire-bonding, and designing and fabricating a custom chip-holder, identifying a number of advantages and disadvantages to each method along the way. Extended thin-film equipment downtime required us to re-focus our work and meant it was not possible to fabricate enough chips to yield a working immunoassay. Instead we designed and tested a low-cost potentiostat, an important tool in not only this project, but in any application that requires electrochemical work. We think this may be of commercial interest and will continue developing this. As far as possible, we completed all the work items required to implement an electrochemical immunoassay as summarised below:1. We initially designed the chips to connect to the readout equipment via a microSD card connector (see Figure 1). Whilst convenient, the long traces and consequent stress buildup led to delamination in liquids. We consequently redesigned the chip and successfully fabricated chips with recessed ring-disc electrode arrays (see Figures 2 and 3) varying from 1-3 µm in diameter and with electrode separations varying from 200 to 400 nm between ring and disc electrodes. Eight to 25 individually addressable pixels varying from 250micrometers to 1 mm in diameter were fabricated using both gold and platinum as electrode materials. As the design has in excess of ten lithography steps and all fabrication steps were carried out in an academic laboratory, numerous optimisation and process verification steps were required. Figure 4 shows an image of an eight pixel chip and a schematic of the layout of the various electrodes.Several etch chemistries and hardmask materials were tested and optimised to facilitate Pt dry etching with thin resists, a process which is especially challenging due to the inertness of the electrode material, but crucial to its function on the chip. This had not previously been done by the Fellow or at NTNU and only became feasible with the purchase of dedicated equipment in the form of an ion beam etcher. A combination of physical (ion beam) etching followed by reactive ion etching produced a higher yield of functional pixels than pure ion beam etching, which resulted in an electrical connection between top and bottom electrodes. Chemical loading effects during the reactive ion etch require optimisation for each hole size. In the continuation of this project, a deep UV stepper will be used to fabricate arrays down to 200 nm – we have designed the reticles, resolved the necessary fabrication challenges and are ready to complete this device development step as soon as we regain access to the necessary etch tools. 2. Reduced delamination of perforated chip layers in solution by changing chip design to reduce stress at corners and developing a protocol to reduce stress in metal and PECVD deposited insulator layers by optimising deposition conditions. A working protocol was developed by a master student supervised by the MC Fellow (see Figure 5). 3. Tested five different methods to interconnect the chips to the electrochemical readout instrumentation: a. flip-chip room temperature bonding of chips with indium/gallium contacts to solder bumped PCBs, followed by epoxy underfill to isolate contacts from electrolyte. Cumbersome and slow process, resulting in many failed interconnects.b. flip-chip room temperature bonding to solder bumped PCBs using anisotropically conducting tape. Could not produce reliable contacts.c. flip-chip bonding of chips with electroplated contacts (5 µm nickel/500 nm gold) to Pb67Sn33 solderbumped and rosin fluxed PCBs at 215degC followed by underfill with epoxy. Most reliable and reproducible contacts, chips self-align. Electroplating adds additional complexity. Not compatible with wafer-scale biofunctionalisation, due to high temperature, functionalisation of chips must be carried out after reflow (see Figure 6).d. room temperature ultrasonic wire-bonding using gold wire to Pt contact pads. Could not get sufficient adhesion of wire to bond pad to produce reliable connections. Although thickening the contact pads with gold to facilitate adhesion is an option, this additional step also opens for flip-chip bonding, which is faster and allows for more facile electrical isolation through underfill with epoxy.e. Custom designed chip-holder machined in PEEK, using spring loaded pogo pins for interconnect to chip contacts and o-ring seal to isolate contacts from electrolyte (see Figure 7). Compatible with wafer-scale biofunctionalisation, and does not require electroplated contacts.2. Designed, built and tested a low cost potentiostat using the commercially available LMP91000 gas sensor analog front end from Texas Instruments with third year student Tomas Hadamek. This €5 chip was interfaced to an Atmel microprocessor for a total cost of less than €35, and connected to a Raspberry Pi single board computer with a small TFT screen to produce a fully integrated potentiostat capable of cyclic, square wave and linear voltammetry experiments for a cost of goods of less than €100 (see Figure 8). Whilst the voltage step size is currently limited to a minimum 0.03V/step like for like results compare favourably to those obtained with a commercial instrument that costs 100 times as much (see Figure 9). This is a promising approach we will continue to follow up on, as a low-cost potentiostat is not only crucial for the future availability of our chip-based sensor platform, but also of commercial interest to a number of other areas including environmental monitoring, battery testing, and science education.3. Electrochemically characterised the chips using generation collection mode, confirming the chips behave as expected, if electrochemically cleaned by cycling in dilute H2SO4 prior to taking measurements (see Figure 10). Without this additional cleaning step, electrode kinetics are poor and do not reach steady state at higher potentials (see attached Masters thesis). Due to the extended equipment downtime at the NanoLab, we had to resort to using chips that had previously been discarded due to electrical shorts between the top and bottom electrode layers. These shorts originated during an attempt to reduce the number of process steps required to etch nanowell ring-disc electrode arrays. Moving to an ion beam only etch would have removed the need to optimise the process for each nanowell diameter (to take account of chemical loading during the reactive ion etch step of the insulating layer). However, redeposition from the top electrode onto the sidewalls of the wells during the ion beam etch was found to short the top electrode to all bottom electrodes (see Figure 11). Several methods were tested to remove the redeposited layer from the sidewalls, including electrochemical etching, resistive heating and wet etching. Whilst none of these approaches worked for the Pt electrode based wafers, the shorts on one test wafer fabricated with gold electrodes were successfully removed using a dilute potassium iodide etch, leading to recessed ring-ring electrode arrays rather than recessed ring-disc electrode arrays.4. A modular immobilisation protocol to attach biotin functionalised probe molecules to amino-silanised silicon nitride surfaces was developed as part of a Master’s Thesis (see Figure 12). A commercial blocking solution (Starting Block from ThermoFisher) was used to prevent non-specific adsorption. The functionalisation and non-specific blocking protocols were characterised using fluorescently labelled streptavidin (see Figure 13). We identified a set of commercially available antibody kits for biofunctionalisation of the chips. These are essential work items to successfully complete the electrochemical immunoassays which are the main goal of this project.5. Investigated two methods for electrochemical amplification (not yet complete).Expected final results and their potential impact and useWe expect to combine the state-of-the-art nanoelectrode chips we have fabricated with novel immunoassay techniques to develop a low-cost, multiplexed, quantitative and ultrahigh sensitivity biosensor platform that will initially be used for immunoassays, but can also be used for other affinity based recognition events such as DNA and aptamer/protein binding. Single board computers, the ease of programming microprocessors, and the availability of analog front-end chips such as the LMP91000 allow the development of hand-held electrochemical instrumentation suitable for research as well as consumer use at a fraction of the cost of currently available systems.Apart from the potential commercial and societal benefits of such a low-cost bio-sensing technology, the chips and potentiostat that have now been produced can already be used as electrochemical sensors in research and teaching, with improved performance and ease of handling compared to a variety of other electrode configurations. We are continuing to work on the technology with an aim towards commercialisation.Contact Details: Dr. Peter Köllensperger – p.kollensperger@ntnu.no final1-figure-10.png final1-figure-6.png final1-figure-8.png final1-figure-13.png final1-figure-1.png final1-figure-7.png final1-figure-2.png final1-figure-3.png final1-figure-5.png final1-figure-12.png final1-figure-11.png final1-figure-9.png final1-figure-4.png
