Understanding the interactions between molecules in complex systems, such as the cells of living organisms, remains one of the main challenges for future biophysical research. Facing the huge number of molecules in living systems, sophisticated detection techniques have to be applied. In this context, the specific and sensitive detection of molecular interactions at solid/liquid interfaces remains one of the most promising approaches. This detection is also becoming increasingly important for a wide variety of applications and technologies, ranging from the analysis of industrial waste to biomedical applications. The main requirement for new sensor technologies is the highly sensitive and specific detection of molecules. However, at the same time, for the applicability of the technique towards screening of biomolecular interactions, the potential of simultaneous parallel detection is mandatory. Surface near, two dimensional electron systems in semiconducting materials are thought to fulfill both of these requirements, a high sensitivity against surface events and the possibility for highly parallelized devices through standard semiconductor technology. In this thesis, a sensor device based on semiconducting Silicon-On-Insulator (SOI) for the detection of biological and chemical molecules has been developed and specific protein binding events to receptors incorporated into lipid membranes have been detected with high sensitivty by these sensors. The SOI wafers used to produce the sensor devices consisted of a 30 nm thick, slightly boron-doped, single crystalline Si (100) sensor layer on top of a buried oxide layer of 100 nm or 200 nm thickness. The substrate wafer underneath the buried oxide was also slightly p-doped silicon and had a thickness of 650µm. The whole structure was covered by a thin (1-2 nm) layer of native oxide. A voltage applied at the back side of the wafer was used to control the charge carrier concentration in the conducting sensor layer and to set the working point for the sensor measurements. A Ag/AgCl reference electrode immersed into an electrolyte on top of the sensor surface was used to control the potential of the electrolyte and to apply front gate voltages. Calibration measurements were carried out to relate the measured changes in sheet resistance to changes in the surface potential. To verify the sensor mechanism, the response of the sensors with native oxide surface against variations of the pH and the salt concentration of an electrolyte was measured and compared to the theoretical predictions of the site-binding theory and the Grahame equation, yielding excellent agreement. The sensitivity of the pH measurement was -50mV/pH and from the noise level, the sensitivity of the developed sensor device was extrapolated to be 0.02-0.04 pH steps. By detecting the adsorption of charged poly-l-lysine, the sensitivity of the sensors towards changes of the surface charge density was estimated to be 1e-/100nm^2. To reduce the drift of the signal over long times and to further functionalize the sensors, the surface was hydrophobized by the adsorption of Poly-Methyl-MethAcrylate (PMMA) or the covalent binding of OctadecylTriMethoxySilane (ODTMS). Both methods yielded a hydrophobic surface as indicated by an increase in the contact angle. For the PMMA covered devices, the pH response was drastically reduced, but the layer was not stable enough for the further use as passivation layer in the lipid membrane experiments. Contrary, the pH response of the ODTMS covered devices did not disappear due to the permeability of the layer for $H^+$ ions, but the layer was very stable due to the covalent bond to the surface. The thickness of the ODTMS layers was determined by ellipsometry to be approx. 1.5 nm, indicating a monolayer of silane on the surface. As biomimetic test system, a functional lipid monolayer with incorporated DOGS-NTA lipids was deposited on the ODTMS covered sensors. The headgroups of the DOGS-NTA lipids were charged and decharged by nickel or EDTA containing buffer and this charging was detected by the sensor with a sensitivity of 1 e-/650 nm^2. A theoretical model within the standard linearized Poisson-Boltzman theory was developed to explain the measured linear increase of the signal with the concentration of incorporated functional lipids. The potential of the sensor for the quantitative detection of biomolecular interactions was demonstrated by measuring the specific and reversible binding of histidine tagged artificial peptides and natural proteins to the functional lipid mono layer. As example petides, the binding of His6Asp6, Green Fluorescent Protein (GFP) and Histidine360LumazinSynthase (His360LuSy) was detected and the obtained signals were quantitatively explained with the previously developed model. The binding of GFP was observed with a sensitivity of 1 molecule/65nm^2 and was shown to be reversible and specific. So far, these binding measurements represent the first quantitative detection of specific protein binding events to a functional and biocompatible lipid layer with a semiconductor sensor device.
«
Understanding the interactions between molecules in complex systems, such as the cells of living organisms, remains one of the main challenges for future biophysical research. Facing the huge number of molecules in living systems, sophisticated detection techniques have to be applied. In this context, the specific and sensitive detection of molecular interactions at solid/liquid interfaces remains one of the most promising approaches. This detection is also becoming increasingly important for a...
»