2.4 Electrolytic Interfaces

A typical workflow for detecting a certain Deoxyribonucleic Acid (DNA) complex is like this: At first the concentration of the DNA sample has to be increased either by Reverse Transcription (RT) or Polymerase Chain Reaction (PCR). Then there would be a step to mark the DNA with a so called label, enabling the detection of the DNA via radiation or light, followed by applying the sample to a microarray. Microarraies contain an array of spots. Every single spot exhibits a different chemical reaction and is therefore able to detect a different type of molecule. After the reaction took place the array is read by an expensive microarray reader.

State of the art technologies to detect pathogens, antigen-antibody complexes, and tumor markers are timeconsuming, complex and expensive [148,149].

Thus, the field-effect transistor comes into play. By replacing the optical sensing mechanism with an electrical signal detection several benefits arise. Firstly, the expensive readout device becomes superfluous. Additionally, using field-effect transistors allows to integrate analyzing and amplifying circuits on the same chip, enabling a further cost reduction due to cheaper equipment. The exceptional development of semiconductor process industry enables mass production of such devices in conjunction with reducing the price per device dramatically. Reams of reaction pairs are attainable and studied: for instance, detecting DNA [17,150,151], cancer markers [152], proteins like biotin-streptavidin [153,154,155,156], albumin [157], and transferrin [158]. These papers show the diversity of device types and materials which where investigated. Even though much promising research has been carried out about BioFETs, they are still in their commencement and many unresolved questions remain.

The BioFET concept is extremely powerful. Literally, every molecule which exhibits a charge within the solute and can be bound to the surface layer of a BioFET can be detected. The devices become smaller, cheaper, and easier to use, so it could enable a family doctor to screen for diseases on his own and decide faster which treatment is the best for a patient, choosing the medication that offers the best results depending on the patients genetic profile. This is one aspect of the so called Point Of Care (POC) applications, which aim to offer help right there where it is needed without much effort. A typically POC application, which is commonly used today, is the blood sugar measurement. The devices are affordable, small, easy to use, and can be applied by the patient. A tiny drop of blood is enough to check the blood sugar level and enables the patient to decide how to proceed. There is also the vision to control the spread of diseases within a population or monitor environmental pollution with such devices.

The effort to integrate a BioFET into a chip environment is not very high. Either by isolating the surrounding areas by a thick oxide or polymere or by putting a microfluidic channel above the functionalized gate of the BioFET the chip can be turned into a mini-laboratory also known as Lab On Chip (LOC). This concept improves the control over environmental parameters like local pH or detecting the amount of a special protein, and it facilitates local measurements (e.g. how a cell reacts to stimulus), thus offering a complete LOC. However, there are still many obstacles to circumvent and a lot more of research is needed.

Park et al. [159] presented a CMOS compatible albumin FET-type sensor for diagnosing nephritis. Nephritis is the inflammation of a kidney and caused by an infection or an autoimmune process. Due to the infection the microscopic filters in a kidney are damaged or closed. This leads to an extraction of important proteins from the blood. Therefore, the characteristic symptoms of a nephritis are called proteinurea. Albumin was identified as a potential candidate to diagnose nephritis. The regular device represents a differential setup and consists of a BioFET able to detect albumin and a reference transistor measuring the background noise. The channel region measured $100\,\mu\mathrm{m}$ in length and $20\,\mu\mathrm{m}$ in width. In human urine an albumin concentration of $<10\,\mathrm{mg}/\mathrm{l}$ is considered as normal, while a concentration above $30\,\mathrm{mg}/\mathrm{l}$ in urine is an indicator of nephritis. The device shows a linear response from $30\,\mathrm{mg}/\mathrm{l}$ to $100\,\mathrm{mg}/\mathrm{l}$. This competes with the established method (urinary stripes), which detects albumin above $100\,\mathrm{mg}/\mathrm{l}$ really good.

Fritz et al. [150] showed a DNAFET with a Poly-L-Lysine (PLL) immobilization layer able to distinct a single mismatch in a $12$-mer oligonucleotide. An oligonucleotide is a short nucleic acid polymer, typically with twenty or fewer bases. They used a differential setup with two sensors in parallel to surpress thermal fluctuations, drift, and non-specific binding. Each sensor exhibited a $50\times50-\mu\mathrm{m}^{2}$ sensing area and was prepared with a different type of oligonucleotide (A and B differ with one mismatch). Injecting complementary to A (cA) into the system only the FET functionalized with oligonucleotide A was able to hybridize and become more negatively charged compared to the FET with oligonucleotide B which could not hybridize. Then complementary to B (cB) was injected and only the FET with B was able to bind the oligonucleotides. Thus, the difference in surface charge vanished again and the differential signal faded out.