University of Patras
Τεχνικές και κυκλώματα βελτίωσης της ανιχνευτικής δυνατότητας αισθητήρων: μοντελοποίηση θορυβικών διαταραχών σε βιοαισθητήρες για διαγνωστικούς σκοπούς
Abstract
dc:descriptionThe need to develop readout circuits that are capable of measuring accurately transducers that are incorporated in various instruments has risen considerably the last few years. These readout circuits cover a broad range of applications and are integrated in a number of different appliances such as smartphones, home network sensors, industrial sensors, portable laboratory equipment, automotive sensors, space mission equipment and other measurement instruments. The present thesis is about sensing devices and experiments from the fields of biochemical sensors and photodetectors of cosmic rays. There is a great demand for the development of fast, reliable and low-cost sensing, monitoring and diagnostic systems of biological molecules and diseases. The next generation of biological detection platforms requires important improvements to the sensitivity and accuracy of measurement, in order to cover the harsh specifications of modern science. The majority of the traditional measurement systems is slow and requires large quantities of sample material to guarantee a reliable diagnosis. Nanotechnology plays an important role in the fabrication of these biosensors. The sensitivity and performance of these sensors is enhanced through the use of nano-materials, employing new signal transduction methods. Due to their small dimensions (comparable with the sample’s target molecules), nano-dimension biosensors allow the fast analysis of multiple substances, both in vitro and in vivo. In the field of high-energy physics, in order to understand the nature and the properties of elemental particles, the use of detection systems is as critical as that of accelerating systems and the related infrastructure. Not surprisingly, modern detection systems become more and more complex, aiming to guarantee the maximum information from each experiment. In nuclear physics, as well as in high-energy physics, the type of detector used depends on the nature of radiation that the reacting particles emit during the experiment. The high accuracy and the capability to compute the detected particles’ energy offers the means to maximize the signal to noise ratio and therefore, increase the telescope’s likelihood of detecting astro-particles. Due to the nature of the detection experiments, the procurement of a reliable measurement is extremely important. Through the study of the transduction process, the mechanisms that lead to the detected signal’s fluctuation become clearer. Affinity based sensors use as a detection structure the BioFET, in order to evaluate the concentration of an analyte molecule in a given sample. Sensors of this type make use of biological probe-molecules, like DNA nucleotides or antibodies, to which the analyte’s molecules are selectively bound. The bound molecules are those that lead to detection. The detection method (transduction process)is critical and differs based on the various biological platform. In this stage, the signal is transformed from “biological” to a measurable, usually electrical, signal. The readout circuit is used to measure this electrical signal, by amplifying and processing it and give as an output the verdict of the presence or not of the target analyte in the sample. The capacitive biosensors exhibit a shift in their electrical capacitance as a response to a natural triggering. The operating principal of capacitive biosensors is based on the fact that the target biomolecule interactions exert a force and deflect the flexible silicon membrane that forms one of the plates of a variable capacitor. As a result of this deflection, the capacitance between the membrane and the silicon substrate changes. In Si NW FETs, for the generation of the respective electrical signals that correspond to the detection of a specific analyte, the target molecules first have to collide with the Si NW surface, on which the receptors are immobilized and participate in a field-effect generation process (basic FET operation). The binding of the analyte’s molecules to the probes results in an abundance of negative or positive charge close to the insulator of the gate, depending on the pH of the sample. This charge causes an equal and complementary charge to appear on the semiconductor surface, under the insulator, via field-effect. This charge influences the conductivity of the semiconductor. In both of these types of biological sensors the important common characteristic is their silicon nature, a feature which makes them readily integrable in any CMOS measurement device. The sensors in the case of particle detection experiments are, in essence, photo-multipliers, which produce as an output electric pulse, whenever an event is detected. Via the measurement of the pulses’ duration, it is possible calculate the particles’ energy, by first reconstructing their waveforms. For the framework of this Ph.D. thesis, various readout circuit architectures were developed, aiming in providing a measurement with improved accuracy compared to the literature. Regarding the particle detection experiments, a readout circuit was designed and fabricated, making use of different ICs. Through the proposed design, an improvement in the detection accuracy is achieved, by using multiple thresholds instead of just one, for reconstructing the particle waveforms. The input pulse is detected by the system and driven to a preamplifier. The preamplifier amplifies this pulse with a fixed gain. The preamplifiers output is connected to three comparators, for each input channel. In the system, three thresholds are chosen for the determination of the duration the pulse spends above these threshold values. This duration is subsequently digitized via a Time-to-Digital converter. In addition, an FPGA is used for the filtering of the data (based on noise and the coincidence of the pulses), the synchronization of the GPS and the communication with the user interface. The system was checked for its proper operation with pulses produced by an actual cosmic ray telescope detector, used to detect extensive atmospheric showers and its resolution was calculated. Regarding the case of capacitive biosensor readout, the developed architecture was implemented in a 90nm IC, and provides a constant resolution for a wide range of input capacitances. The analog part uses a switched-capacitor amplifier to amplify a reference voltage. Since the feedback capacitance is a known constant and the input capacitance is the sensor capacitance to be measured, the gain of this stage varies, based on the biosensor’s value. An ADC measures the output voltage of the amplifier to extract the required information. A finite state machine developed in FPGA logic is used for the formation of the reference capacitance, from a bank of external capacitors. The novel architecture essentially implements a form of Correlated Double Sampling (CDS) and therefore exhibits reduced noise at low frequencies. After measurements the accuracy and the linearity of the system are evaluated, the measurement uncertainty is analyzed and the measurement error is computed. Furthermore, the noise behavior of Si NW FETs is studied, since they are an advanced version of biological sensors. The mechanisms that lead to gate voltage fluctuations are studied, in order to extract information relevant to the biological reaction, using noise spectroscopy. The noise spectrum of Si NW FETs includes all the information for the transport phenomena, the molecular interactions of binding/unbinding and the reaction rate of the experiment. This method of information extraction, noise spectroscopy, is very useful not only for the further improvement of the sensor device, but also for using noise itself as a detection mechanism. The biochemical reaction that takes place during a detection experiment with a Si NW FET sensor is related to two types of fluctuations, namely the binding/unbinding of the reacting molecules and also, the binded molecule’s charge fluctuations during thermal equilibrium. These fluctuations have an immediate effect to voltage fluctuations on the gate, which in turn affect the electrical characteristics of the sensor device, via field-effect action. The resulting noise contains information about the detection, which is extracted via the time constants (characteristic frequencies) that are related to the kinetic behavior and the charge fluctuations of the target molecules. A charge fluctuation model is developed to describe the fluctuations on the sensor surface, which are linked to the two above mechanisms. The spectral density corresponding to these types of fluctuations is added to the overall noise spectrum of the device. From the noise spectrum analysis the characteristic frequencies stemming from a given target analyte are extracted. From these two noise mechanisms, the characteristic frequency of the charge fluctuations is readily visible in the overall noise spectrum, since it appears in the kHz range. The dependence of this characteristic frequency on the receptor molecule concentration leads to the possibility of detecting multiple analytes concurrently, from the same Si NW matrix. To that end, the fabrication of a Si NW matrix is proposed, with high resolution and receptors of different, a priori known densities. The spectral density corresponding to a certain type of receptor has its own unique characteristic frequency in the overall noise spectrum. In addition, a decomposition algorithm is implemented for the extraction of the unique characteristic frequencies from the spectrum, in order to achieve the determination of the number and type of target molecules that are present in the electrolyte sample.
Degree
thesis:*- Grantor dc:publisher
- University of Patras
- Year dc:date
- 2017
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Georgakopoulou, Konstantina
- Γεωργακοπούλου, Κωνσταντίνα
Subjects
dc:subject × 10Rights
- Language dc:language
- gre
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/41832