DE10254158A1 - Sensor system for detecting signals from biological cells comprises semiconductor substrate containing field effect transistor whose channel has coupling component connecting it to sensor surface - Google Patents

Abstract

Sensor system for detecting signals from biological cells (24) comprises a semiconductor substrate (SUB) with an upper surface (26) on which the cells are placed. The substrate contains a field effect transistor whose channel has a coupling component connecting it to a sensor surface (16). Sensor system for detecting signals from biological cells (24) comprises a semiconductor substrate (SUB) with an upper surface (26) on which the cells are placed. The substrate contains a field effect transistor whose channel has a coupling component connecting it to a sensor surface (16). The normal to the sensor surface is at an angle to the normal to the upper surface of the substrate. The coupling component allows the conductivity of the channel to be affected by signals from the cells, using capacitive coupling. An Independent claim is included for a method for detecting electrical signals from biological cells using the sensor system

Description

The invention relates to a sensor device for Detection of electrical signals from a biological cell and a Method for the detection of electrical signals of a biological Cell using an appropriate sensor device.

The detection of electrical potentials in biological cells by means of extracellular detection beginnings already so advanced that you now have one growing commercial importance, for example in the area of high-throughput screening in the pharmaceutical industry. One reason for this development is in the non-invasive contacting, which is very good at artifact-free Measurements and for the long-term detection of such signals. As Sensor devices are preferably used on planar substrates their surface arranged a plurality of metal electrodes in a field-like manner is. With these sensors different biological systems observed, e.g. organotypic brain sections or single cell cultures. A sensor surface of a Such sensor device usually guides the sum of the Signals from several cells simultaneously (so-called multi-unit recording).

The derivation or detection of the electrical signals of the biological cells is based on a capacitive Coupling the extracellular sensor surface with the cell, for which a narrow spatial Near the sensor surface to the measurement object needed becomes.

For an efficient capacitive coupling should be the distance between the sensor surface and Cell membrane preferably be less than 40 nm. The capacitive Measurement Method allows the construction of spatially dense, but still crosstalk-free Sensor arrays.

In the publication by S. Vassanelli and P. Fromherz, "Neurons from Rat Brain coupled to Transistors ", Applied Physics A 65, pages 85 to 88 (1997) was shown not to only with invertebrate cells, but also with the much smaller ones Vertebrate cells a single cell derivation (so-called single unit recording) with extracellular Electrodes is feasible. allows this was achieved through the use of field effect transistors (FET) active sensor elements. The use of such FET-based Sensor devices that are complete Made in CMOS technology overcomes major disadvantages passive multi-electrode arrays, in particular through signal amplification close to the sensor, signal processing and line number reduction.

Both passive, metal electrode-based active, FET-based sensor devices are also becoming typical structured in layers using photolithographic methods and therefore naturally have a planar sensor surface on. In the cultivation of neural networks from single cell supervisory favor the flat surfaces Conflict between the neuronal growth dynamics and the fixed ones Sensor positions, causing significant problems with single cell signal derivation leads. For example, in the dissertation by M. Jenkner, "Hybrid Networks from neurons of Lymnaea stagnalis and silicon chips ", Technical University of Munich (1999), that vertebrate Single cells in culture change their position within a few days Multiples of the sensor surface diameter change can and therefore soon away from the detectable sensor area extracellular sensor device, on which they originally were placed.

The neuronal growth dynamic is a well-known biological phenomenon, the vital cellular Underlying mechanisms. A typical neuronal appearance Growth dynamics can be seen in single cell cultures, which in vitro again grow together into neural networks. Are the neurites two adjacent? Neurons have come into contact, so the resulting neuritic Fibers the tendency to shorten. The forces involved are so big that classic adhesives are unable to return the Somata to its original Hold position. However, the success of extracellular signal derivation is because of the short-range capacitive coupling strongly from the right one Position of the nerve cell to be detected depends.

Therefore, in the prior art a variety of methods have been proposed, such as the adverse influence of neural Growth dynamics on conventional ones capacitive sensor device can be reduced.

According to a solution known as the "scrap shot method", sensor devices are used which use highly integrated sensor FETs in large, dense fields for extracellular signal derivation. This opens up the possibility that, on the one hand, the individual neurons do not have to be individually placed on the sensor surfaces when the culture starts. On the other hand, if the sensor fields are sufficiently dense, it is ensured that a neuron which leaves a sensor area due to the growth dynamics will again lie on an adjacent sensor area in the immediate vicinity. This approach is examined in numerous current projects. However, it is unclear whether the vertical distance to the sensor surface does not increase during the displacement of the neuron and thus weakens the signal transmission. The desired contact probability of a neuron is guaranteed with the scrap shot method due to the extremely high sensor surface density and number. However, the high number of sensors requires a drastic reduction in the number of lines by multiplexing with the associated build-in the signal-to-noise ratio.

In other conventional devices additionally to the sensor area "cages" used to control the mobility of the "trapped" neuron. In both cases the cell soma is fixed, but the neurites are given the opportunity to out of the cage grow out.

For example, a suspension became young Rat cells cast on a sensor device, the sensor surfaces on Lie at the bottom of a depression. The opening of the depression will overlapped at the corners by micro-mechanical struts, so that a partially closed cage with defined opening results. The size of the opening was set exactly so that only very young cells fit through. Due to the slight growth of the cell soma after a few days, only the neurites grow out of the partial cage. The grown Soma, however, was in the partial cage locked and thus coupled to the sensor surface. However, can buried this process with " Cages "only for very young Cells are used and is inflexible in terms of cell cultures with greater dispersion the somatic diameter. In addition, the production requires complex processes, as used in MEMS.

A more advantageous sensor device arrangement and a simplified production was described by G. Zeck and P. Fromherz in "Noninvasive neuroelectronic interfacing with synaptically connected snail neurons immobilized on a semiconductor chip", Proceedings of the National Academy of Sciences, 98, 18, pages 10457 to 10462 ( 2001). The sensor areas of a conventional FET-based multi-electrode array were surrounded by a ring of polyimide columns which, similar to the previously described method with "buried cages", fixed the cell soma but did not hinder the growth of neurites. In 9 is such a sensor device in schematic cross section ( 9 (a) ), in a schematic perspective view ( 9 (b) ) as well as light microscopic supervision ( 9 (c) ). The cell 100 comes here on a contact surface 102 to lie, which coincides with the sensor surface of the sensor. The investment area 102 is formed by the surface of a gate oxide layer which is arranged on a channel region of a planar MOSFET. From the cell 100 Electrical signals generated "couple" capacitively into the sensor transistor via the field effect, which results in conductivity modulation of the conductivity of the channel region KANAL between the source region SOURCE and drain region DRAIN. The soma of the cell 100 is in its movement parallel to the substrate surface plane through the cage-like polyimide columns 104 so limited that it covers the sensor area 102 can't leave. In 9 (b) are only two of the six polyimide columns for better clarity 104 (see. 9 (c) ). With this arrangement, in contrast to the latter method, soma and neurites remain in one plane.

Given the problems mentioned above It is an object of the invention to provide a sensor device for detection or deriving electrical signals from a biological cell and a method for the detection of electrical signals of a biological Cell.

This task is accomplished by a sensor device with the features specified in claim 1 or a method for Detection with the features specified in claim 15 solved. preferred embodiments are subject to the dependent Expectations.

• – ein sich entlang einer Substratflächenebene erstreckendes flächiges Halbleitersubstrat mit einer An- bzw. Auflagefläche, an welche die biologische Zelle anlegbar ist;
• – zumindest einen auf dem Halbleitersubstrat ausgebildeten
• Feldeffekttransistor, welcher einen Source-, einen Drainund einen dazwischen angeordneten Kanalbereich aufweist;
• – zumindest ein an den Kanalbereich angrenzendes Kopplungselement mit einer Sensorfläche, wobei eine Normalenrichtung der Sensorfläche zumindest bereichsweise gegenüber einer Normalenrichtung der Substratflächenebene geneigt ist, so daß die Sensorfläche von der Anlagefläche vorspringt;

wobei das Kopplungselement derart ausgelegt ist, daß die elektrischen Signale der an die Sensorfläche angrenzenden biologischen Zelle über eine kapazitive Kopplung die elektrische Leitfähigkeit des Kanalbereichs beeinflussen.According to the invention, a sensor device for detecting or deriving electrical signals comprises at least one biological cell

• A flat semiconductor substrate extending along a plane of the substrate surface and having a contact or support surface to which the biological cell can be applied;
• - At least one formed on the semiconductor substrate
• Field-effect transistor, which has a source, a drain and an intermediate channel region;
• - At least one coupling element adjoining the channel region with a sensor surface, a normal direction of the sensor surface being inclined at least in regions relative to a normal direction of the substrate surface plane, so that the sensor surface projects from the contact surface;

wherein the coupling element is designed such that the electrical signals of the biological cell adjacent to the sensor surface influence the electrical conductivity of the channel region via a capacitive coupling.

The sensor device according to the invention is an active, FET-based detection structure. The electrical signals of the biological cell, in particular of a neuron, are detected by modulating the channel conductivity of the channel of a (sensor) field-effect transistor. For this purpose, the sensor device has a field effect transistor which interacts with a coupling element. The coupling element is arranged in particular on the channel region, that is to say that semiconductor region in which the transistor channel is to be formed. The coupling element has a sensor surface, in the immediate vicinity of which the biological cell is brought to the detection or derivation of its electrical signals. The electrical potential variations generated by the cell are capacitively coupled into the channel region via the coupling element, as a result of which the electrical channel conductivity is influenced due to the field effect depending on the electrical signals. The cell thus acts similarly to an electrically conductive gate of a conventional MOSFET.

The sensor device according to the invention is for Single cell recording (single unit recording) designed so that electrical Signals from a single biological cell are detectable. however can in a known manner with a large number of field effect transistors cooperating coupling elements may be provided so that at the same time the electrical signals of a large number of biological cells are independent of one another can be examined (single cell derivation of a large number of cells).

The sensor device is based on a flat semiconductor substrate extending along a plane of the substrate surface, especially a silicon wafer, on or on which a contact or support surface is provided for the biological cell. The contact surface can be directly on the semiconductor substrate or on an or several layers attached to it may be provided. The contact surface provides a surface represents the sensor device to which the biological test to be examined is operational Cell approximated becomes. The cell must do this do not come into direct physical contact with the contact surface. Rather, you can choose between contact surface and there is a certain distance between the cell and the electrolyte solution (nutrient solution), in which the cell is typically suspended.

In the sensor device according to the invention fall differently than conventional ones Sensor devices contact surface and sensor surface not together. Instead, the coupling element has a sensor surface, the normal direction of which, at least in regions, with respect to a normal direction the substrate surface plane (Wafer or chip level) is inclined, i.e. one different from 0 ° Forms an angle with it. The sensor surface is opposite the Substrate surface plane inclined and arranged so that it protrudes from the contact surface. Between sensor surface and investment area thus one different from 0 ° Angle formed. This causes the protruding from the contact surface sensor surface additionally the sensor function also fulfills a holding function for the biological cell. While at usual Sensor devices the contact surface with the sensor surface coincides and the guaranteed with it Signal transfer sensitive to any change in position due to capacitive coupling the cell responds, are the two in the inventive sensor device surfaces separated from each other. The projecting sensor surface thus shows a difference to conventional Sensor devices a the mobility of the cell in the substrate surface plane restrictive hold function on, which creates a closer and more stable contact between cell and sensor surface becomes possible. Due to the distance sensitive capacitive coupling between Cell wall and sensor area such a position stabilization of the coupling element the cell has a signal-improving effect.

Ideally, the cell borders if possible directly to the sensor surface to be as efficient as possible to effect capacitive coupling. However, in reality, between the cell outer wall and the sensor surface there is practically always an unavoidable little space left which is typically filled in by the nutrient solution in which the cell is suspended. Likewise located between the contact surface, which represents an adhesive surface, and the cell outer wall practically always a small gap, which is typically caused by Nutrient solution is filled.

Compared to the conventional "scrap shot method", a sensor device according to the invention has significant advantages. So with the conventional No topographical help to stabilize the process Neuron-sensor surface contact given what pays with high circuit complexity and poorer signal-to-noise ratio must become. In contrast to the known methods described at the beginning with "buried and raised Cages "are sensor surface and Holding device separated from each other, increasing the width of use is significantly expanded. In contrast to the sensor device according to the invention have to in the known methods the diameter of the cages is very great can be adapted exactly to the diameter of the neuron to be examined.

The coupling element preferably consists of at least one dielectric layer. The coupling element is thus preferably formed from a single or multilayer structure of dielectric layers which adjoin the channel region of the field effect transistor. The layer thickness of the single-layer or multi-layer structure of the coupling element is preferably essentially constant over the channel region, so that the sensor surface of the coupling element runs parallel to the interface of the channel region. In the case of such an embodiment, a field effect transistor is used which differs from conventional planar transistors in that the interface of the channel region or the sensor surface is inclined with respect to the plane of the substrate surface. In such a coupling element, there are no electrically conductive layers, in particular no metal or cook-doped polysilicon layers, between the sensor surface and the interface of the channel region. Materials such as are used for gate oxides in CMOS technology, in particular oxides, preferably silicon dioxide, and materials with high dielectric constants are particularly preferably used to form the dielectric single-layer or multilayer structure (so-called high-k materials). The cell, the electrical signals of which are to be derived, effectively represents the "gate electrode" of the gate-less structured transistor.

The coupling element preferably has a Layer thickness in the range from 1 nm to 100 nm.

As an alternative to the coupling element, which exclusively consists of dielectric layers, coupling elements can also Find use which have an electrically conductive layer. at this alternative embodiment can for example adjacent to one located at the channel area dielectric layer a thin Metal electrode can be provided, which extends up to a sensor section of the coupling element continues. The sensor surface is, for example, by a surface a dielectric layer is formed, which on the continued metal electrode is arranged. sensor surface and interface of the Can range therefore aligned differently (different normal directions) and be spaced apart. The metal electrode acts as capacitive divider by the electrical signals to be derived first capacitively to the metal electrode and capacitively from the metal electrode couple the channel area of the transistor. Although in this embodiment planar field effect transistors are used, however, suffer due to the lower coupling efficiency of the electrical to be derived Signals to the sensor transistor the signal noise characteristics of the sensor device.

The contact surface is preferably essentially parallel to the substrate surface plane. The contact surface is preferably a planar surface.

According to a particularly preferred embodiment is the normal direction of the sensor surface compared to the normal direction of the Substrate surface plane by an angle in the range of 30 ° to 130 °, preferably about 90 °, inclined. Thus, the sensor surface projecting from the contact surface is preferably present essentially perpendicular to this. The height of the sensor surface, i.e. the distance in the normal direction of the contact surface from this to the upper edge the sensor surface, is preferably dependent chosen from the diameter of the cell and is typically 1 to 100 µm.

According to another particularly preferred embodiment the sensor device has a plurality of the field effect transistors each with one of the coupling elements and the coupling elements are arranged so that the sensor surfaces of the coupling elements at least partially a cell space in directions parallel to the substrate surface plane limit. In this preferred embodiment of a sensor device according to the invention are a large number of field effect transistors with associated coupling elements provided whose sensor surfaces a "cage" for the patient to be examined Form cell. The sensor areas of the coupling elements thus limit parallel to directions the substrate surface plane at least partially a cell space, which is used to hold the investigating cell is designed. The cell can move in directions parallel to the substrate surface plane only move within the cell space and is moving across the Cell space limit through the sensor surfaces, which have a holding function comes, hindered. The size of the cell space can be chosen freely and is typically in the range of 1 to 100 µm.

The sensor surfaces are preferably which limit the cell space laterally, arranged so that between the sensor surfaces Free spaces or crossings are provided, by which the dend thirds of a person to be examined Neurons can grow out of the cell space in the lateral direction. The interspaces are preferably smaller than the cell soma, so that this prevented from leaving the cell space by the sensor surfaces becomes. The cell space is limited at the bottom by the contact surface and is preferably open at the top. However, it is also possible to one of the contact area opposite top edge of the sensor surfaces projections to be provided, which serve as movement barriers for the cell soma in one direction perpendicular to the substrate surface plane Act.

This results in a cage-shaped arrangement field effect transistors with coupling elements attached to them, which lead to an efficient movement inhibition of the cell preferably its inclusion in care for the cell room. In contrast to the conventional ones mentioned at the beginning raised or buried cages can the lateral diameter of the cell space clearly from the diameter of the neuron deviate because the lateral location of the cells does not is identical to the sensor location. The neurite shortening, which in all previous approaches signal transmission between neuron and sensor surface impaired has, carries in such an embodiment even for signal improvement. So it was mentioned in the beginning release of G. Zeck et al. shown that the usually spherical Because of the neuritic shortening pull Somata so strongly on the "cage" or the "fence" made of polyimide columns, that a severe deformation of the soma can occur. The present one in particular preferred embodiment uses the preferably vertical arrangement of the field effect transistors this fact from that the for the extracellular Deriving substantial distance between the cell membrane of the soma and the sensor surface reduced by the neuritic train and the efficiency of the capacitive signal transmission so that is enlarged.

Likewise, in the above publication also shown that several of the limiting polyimide columns in were in close contact with the cell soma. In the proposed here particularly preferred embodiment a sensor device according to the invention results this makes it easy to use several sensor surfaces simultaneously derive electrical signals of the same neuron extracellularly. Depending on size and neuritic Linking the Neurons there is also the possibility on opposite Sides to derive electrical signals from a neuron. Furthermore, increase such redundant contacting or coupling options the contacting probability between sensor and neuron. Are all sensor surfaces for capacitive contacting of the neuron, there is an increase in the active sensor surface and a correspondingly improved signal coupling in and coupling out in the cell or in the transistor.

According to another preferred embodiment the normal directions of the sensor surfaces almost intersect in a central area of the cell room. In other words the sensor areas of the coupling elements with respect their normal directions arranged so that they are all in the direction of Cell space, especially in a central area of the cell space, demonstrate. The diameter of the central area of the cell space is 1/5 to 1/10 of the total diameter of the cell space. The cell room can for example, have a round or a rectangular shape. The normal directions of the sensor surfaces particularly preferably intersect a single point.

According to another particularly preferred embodiment the semiconductor substrate is an SOI substrate (silicon on insulator substrate) and the field effect transistor is a FINFET. Such FINFETs will processed in a web-like rib (fin) made of semiconductor material, which is preferably on a buried oxide layer (Burried Oxide BOX) of the SOI substrate is arranged. The current flow in the channel area of such FINFETs in progress parallel to the substrate surface plane. The channel range of FINFETs can be changed using the field effect with rectangular Channel cross section from three sides (top and the two side surfaces of the Fin) can be varied.

The FINFET is preferably such designed that the on a channel side surface of the channel area arranged coupling element much thinner than Is field dielectrics, which at the other interfaces of the web-like semiconductor rib are arranged that are not connected to the coupling element are provided. This creates a capacitive coupling between Cell wall and transistor channel promoted in the area of the thin coupling element Compared to the much lower coupling due to the thick Field dielectric, which the rest interfaces of the channel area of the FINFET. Preferably, the Channel side surface, to which the coupling element adjoins in a plane of the substrate surface vertical direction. The sensor surface preferably runs parallel to the channel side surface. The channel region can be either n-, p- or undoped.

Preferably extend the Source and drain regions of the FINFET in parallel directions to the normal direction of the sensor surface away from the sensor surface. Thus, the source and drain regions of the field effect transistor across from the sensor surface and opposite of the cell to be examined or turned away so that they a comparatively large one Distance from the cell. This eliminates interference at Avoided cell signaling, which could occur when the source and drain areas with corresponding metallic leads in close spatial contact would come with the cell.

According to another particularly preferred embodiment the channel area has two opposite interfaces, being at one of the interfaces the coupling element and one at the other of the interfaces Backgate electrode or back electrode is arranged. Has particularly preferred in this embodiment the channel area of the FINFET in a section plane perpendicular to the direction of current flow an essentially rectangular one Shape. The result is two parallel mesa or channel side surfaces, whereby on one of the canal side surfaces the coupling element and on the other of the channel side surfaces Backgate electrode is arranged. Using the backgate electrode can a second transistor channel on the opposite of the coupling element Interface of the Channel area are formed. For this, the vertical distance between the two opposite interfaces preferably at least 50 nm, preferably 100 nm. Thus, in the same Channel area (same fin) with sufficient channel or body thickness a second transistor can be formed, which are used can to process-related Compensate for parameter variations of the FINFET.

According to another particularly preferred embodiment the field effect transistor is a vertical transistor (VFET), at which of the source, channel, and drain regions in semiconductor layers are formed, which are in the normal direction of the substrate surface plane adjoin each other. Even with such a vertical transistor concept can easily create an interface of the channel area which are preferably substantially perpendicular to the substrate surface plane or the contact surface is aligned.

All of the source regions of the field effect transistors are preferably formed in the same semiconductor layer as a common source region. So with a vertical transistor concept As described above, if a large number of field effect transistors with associated coupling elements are used to form a "cage-like" arrangement for the cell to be examined, a common source region can be used, which considerably simplifies the electrical contacting of the transistors. To suppress the floating body effect, an ohmic body contact can be formed on the channel region of the FINFET.

There is preferably one on the semiconductor substrate additional Evaluation electronics for the electrical signal to be detected is formed. The evaluation electronics is preferred in conventional CMOS technology using planar (Bulk) MOSFETs created and located in immediate spatial Proximity to the (sensor) field effect transistor. Because of the possible monolithic Integration of sensor and evaluation structures, especially signal amplification elements and switches, can significantly improve the signal-to-noise ratio as well as a higher one Component integration can be achieved.

According to another aspect of Invention becomes a method for the detection of electrical signals proposed at least one biological cell, the following Steps include:

• - Provide a sensor device according to the invention;
• - Suspend the biological cell in a solution;
• - Apply the solution on the sensor device so that one of the biological cells to the contact surface and the sensor surface adjacent; and
• - Detect or deriving the electrical signals of the cell via its Influence on the electrical conductivity of the channel area by measuring the electrical conductivity of the channel area.

Preferably the biological Cells in an electrolytic solution or nutrient solution suspended. For individual signal derivation, a cell is preferably attached to the contact surface and to the sensor surface created, whereby - how described at the beginning - none direct physical contact between cell wall and system or sensor surface necessary is. Rather, there is a (small) space between the cell wall and system or sensor surface remain which is filled in by the solution. The electrical Signals from the cell are detected via their field effect on the electric conductivity of the transistor channel. The fact that the contact surface and the sensor area at the inventive method are separated, with the sensor surface projecting from the contact surface and opposite the substrate surface normal is inclined, one can face conventional detection methods improved cell sensor contact can be achieved. With the sensitive Influence of Sensor surface cell wall distance in the capacitive coupling of the cell soma to the sensor surface this ensures an improved detection signal.

The invention is described below Reference to accompanying drawings of preferred embodiments described as an example. It shows:

1 (a) - (D) schematic perspective views of intermediate stages of a preferred embodiment of a sensor device according to the invention during manufacture;

2 a schematic perspective view of a further preferred embodiment of a sensor device according to the invention;

3 a schematic perspective view of a further preferred embodiment of a sensor device according to the invention, which uses FINFETs;

4 is a schematic perspective view of a modification of the in 3 illustrated embodiment, wherein a FINFET has a back gate electrode;

5 (a) - (C) electrical block diagrams of preferred calibration circuits for the in 4 illustrated embodiment;

6 schematic plan view of a further preferred embodiment of a sensor device according to the invention with four FINFETs with recessed source and drain regions;

7 schematic sectional view of a further preferred embodiment, in which a planar electronic evaluation system is arranged in close proximity to the sensor transistor on the semiconductor substrate;

8 (a) - (F) schematic sectional views of further preferred embodiments of sensor devices according to the invention;

9 (a) is a schematic sectional view of a conventional sensor device;

9 (b) is a schematic perspective view of the in 9 (a) illustrated sensor device; and

9 (c) a light microscopic view of the in 9 (a) Sensor device shown with introduced biological cells.

In 1 (a) - (D) Process stages of the manufacturing process of a preferred embodiment of the sensor device according to the invention are shown. 1 (a) shows a schematic, perspective view of the starting material for the production of the sensor device. A layer sequence of thin semiconductor layers is epitaxially deposited on a semiconductor substrate SUB, which is preferably a single-crystalline silicon substrate. A highly doped semiconductor layer is first applied to the semiconductor substrate SUB 10 , which later forms the common source area SOURCE of the sensor device, an undoped or less doped Semiconductor layer 12 , which later forms the channel area CHANNEL of the field effect transistor and, at the top, a again highly doped semiconductor layer 14 , which later represents drain areas DRAIN of the sensor device. The layers 10 . 12 and 14 extend in the substrate surface plane, which runs parallel to the surface of the semiconductor substrate SUB.

Starting from the semiconductor substrate SUB thus prepared with the epitaxial semiconductor layers 10 . 12 and 14 In subsequent process steps, an arrangement of vertical transistors (VFETs) is structured, which preferably form a cage-shaped arrangement for the biological cell to be examined. The doping profile of the epitaxial semiconductor layers 10 . 12 and 14 can be selected as both a npn and pnp profile. In 1 (a) is also schematically an insulation S. TI (shallow trench isolation) of the active area is indicated by an etching step in the substrate SUB. The layer thickness of the individual epitaxial semiconductor layers 10 . 12 and 14 depends on the size of the biological cell to be fixed later, which typically have a diameter of 10 to 100 μm. The doping profile of the semiconductor layers 10 . 12 and 14 is preferably determined in situ already during the epitaxial deposition. Alternatively, however, an internal implantation can also be used to form the desired doping profile.

The semiconductor layers are subsequently described using a suitable etching mask, for example a resist mask, a TEOS or nitride auxiliary layer (hard mask) 10 . 12 and 14 structured in vertical semiconductor meses and then etched from the epitaxial layer stack using reactive ion etching. The preferred mesa has a substantially vertical mesa side wall, ie a side wall which is arranged perpendicular to the plane of the substrate surface. Equally, however, pyramid-shaped mesas, for example, or concave mesas, for example, can also be produced by special anisotropic etching methods. The depth of the mesa etch depends on the height of the epitaxial semiconductor layers 10 . 12 and 14 , which are preferably silicon layers. The goal is the channel layer 12 safe to etch through and in the (lower) source layer adjacent to the semiconductor substrate SUB 10 to stop. Of course, the drain regions in the (lower) semiconductor layer can be significantly influenced without affecting the operational functionality of the later sensor device 10 and the source regions in the (upper) semiconductor layer 14 be formed. The manufacturing stage after completion of the reactive ion etching step is in 1 (b) shown.

A dielectric layer (not shown) is then grown over the entire surface of the basic structures of the VFETs. The at least one dielectric layer represents the coupling element for the capacitive coupling between the vertical transistor channel and the biological cell and is in particular arranged on the channel region CHANNEL of the vertical transistor. For example, the coupling element consists of a thin oxide layer, for example a silicon dioxide layer, as is used as gate oxide in conventional planar MOSFETs. The coupling element (not shown in detail) borders directly on the interface (mesa side surface) of the exposed silicon mesa structure. The dielectric layer (not shown), whose surface facing away from the channel region CHANNEL is a sensor surface 16 of the vertical transistor has a much smaller layer thickness than field dielectrics, which protect the other interfaces of the channel region. This ensures that the coupling element only in the area of the sensor surface 16 permits an efficient capacitive coupling between the cell and the channel area KANAL, while the capacitive coupling into the channel area KANAL is kept very small over the remaining mesa interfaces due to the thick field dielectric.

In 1 (c) are also source contacts 20 which is the semiconductor layer 10 for (contact SOURCE shared source area). The drain areas DRAIN of the individual vertical transistors can be connected to separate drain contacts 22 connect independently. Not shown in 1 (c) are the wiring of the contacts 20 and 22 , As an alternative to metals, high-density doped polysilicon tracks or alternative highly conductive materials can also be used. The metal tracks (not shown) for contacting the vertical transistors run in the in 1 Embodiment shown above the cages formed by the VFETs. Nevertheless, several such cage arrangements can be combined to form a cell array by leaving space for connecting neurites during processing between the transistor cages.

l (d) shows schematically a biological cell inserted in the "cage" which is formed by the individual vertical transistors 24 , The cell 24 is shown simplified as a spherical structure. The cell room 18 or the biological cell used 24 facing mesa side surfaces represent the sensor surfaces 16 of the vertical sensor transistors. Only through these sensor areas 16 efficient capacitive coupling is possible because the remaining mesa interfaces are protected by much thicker field dielectrics. Between the individual in l (d) The vertical mesa structures shown are laterally provided with spaces or gaps that are smaller than the typical diameter of the cell soma. However, neurites can grow through these spaces and, if a cell field is formed, connect to neurites of other cells. The Neuritenver Shortening, which has impaired the signal transmission between neuron and sensor surface in all conventional sensor devices, even contributes to signal improvement in the present arrangement. Because of the neurite shortening, the essential distance for the capacitive coupling between the cell membrane of the cell soma and the sensor surface becomes 16 reduced, which increases the efficiency of capacitive signal transmission.

Another advantage of in 1 The preferred embodiment shown is the possibility of simultaneously deriving the same neuron extracellularly using a plurality of sensors (a plurality of vertical field effect transistors with coupling elements with sensor areas). Depending on the cell space size and soma size, as well as the neuritic connection of the neuron, there is the possibility of deriving electrical signals from opposite sides of a neuron. The redundant contact possibility and the increased contact probability of the neuron are also advantageous here. All sensor surfaces 16 the vertical transistors used for signal derivation, there is a considerable increase in the active sensor area and a correspondingly improved signal-to-noise ratio due to better signal input and output. Coupling into the cell or into the transistor.

Compared to conventional sensor devices, which, for example, polyimide columns use, the presented preferred embodiment a much simpler manufacturing process, since this is completely with conventional CMOS manufacturing processes is compatible. The vertical transistors do not have to go through special processes, such as micromechanical process steps or non-CMOS materials how polyimide can be enriched, thereby making the sensor device is greatly simplified. Are also combined for the manufacture Sensor holder no additional masks required. Instead of that in the prior art used polyimide etching will now be a silicon etch used. The silicon trench or silicon mesa can do the opposite to polyimide, which forms the holding structure in the prior art, for further Circuit elements (active and passive components) are used. In contrast to the conventional one Sensor device with polyimide cages there is the device according to the invention preferably exclusively from mechanically strong, inorganic materials, what the handling and cleaning of the sensor device much easier. The geometry of the trench (or inverse of the mesa) can be planar via the Layout of the mask used as well as vertically (vertically, conically) over the etching process used (isotropic / anisotropic) be optimized.

In 2 Another preferred embodiment of a sensor device according to the invention is shown in a schematic, perspective view. In the following, the same or similar features of further embodiments of sensor devices according to the invention are provided with the same reference symbols, and a renewed description is dispensed with. At the in 2 Embodiment shown is in the final phase of the manufacturing process, which is similar to that of 1 runs, the surface planarized with an oxide, for example. Below is a trench for the biological cell 24 opened to the cell room 18 define. The source contact 20 is via a contact hole to the semiconductor layer 10 connected. Because of the flatter topology, this embodiment is in contrast to the embodiment according to FIG 1 easier. On the other hand, the embodiment of 1 cheaper for neurite growth.

The source and drain lines of the vertical transistors are supplied by metal tracks that through a sufficiently thick passivation compared to the electrolyte or Nutrient solution, in which the biological cells are suspended can be isolated have to. Therefore, there are embodiments preferred of the transistors in which the metal tracks to the source or drain contact at a clear distance in front of the transistor channel end up. This distance is preferably at least as large as that Passivation layer thickness.

3 shows a schematic, perspective view of a further particularly preferred embodiment of a sensor device according to the invention. So-called FINFETs are used in this embodiment. FINFETs are distinguished by the fact that their current flow, in contrast to the vertical transistors previously presented, runs parallel to the substrate surface plane. The source, channel and drain region of a FINFET transistor is formed in a web-like rib (fin) made of semiconductor material, preferably single-crystal silicon. The starting point for the production is an SOI substrate (silicon-on-insulator substrate), in which a buried oxide layer BOX (buried oxide) is arranged on a semiconductor substrate SUB, on which a thin, single-crystal (TOP) silicon layer is applied. Thin semiconductor ribs or fins are structured in this thin (top) silicon layer, which extends parallel to the substrate surface plane, in which the source region SOURCE, the channel region KANAL and the drain region DRAIN are formed.

At the in 3 In the preferred embodiment shown, the fin made of semiconductor material has an essentially rectangular cross section in a sectional plane perpendicular to the fin longitudinal axis (current flow direction). The lateral boundary surfaces of the semiconductor fin are therefore preferably oriented perpendicular to the plane of the substrate surface. The coupling element, which in 3 is not shown in detail, preferably consists of one or more dielectric layers, which are arranged in particular on the channel region CHANNEL of the FINFET. The sensor area 16 of the coupling element preferably runs parallel to the mesa side surface of the channel region CHANNEL of the FINFET and is thus parallel to a surface normal of the contact surface 26 , The metal line for source contacts 20 and drain contacts 22 come at the in 3 FINFET embodiment shown to lie in one plane. The metallization level can be above or below the support level of the cell 24 are located. The production of such FINFETs with a horizontal channel is based on the known methods for producing FINFETS.

A particularly preferred modification of the in 3 FINFET embodiment shown is a schematic top view in 6 shown. In this embodiment, the source and drain regions are DRAIN, SOURCE from the cell space 18 or from the cell 24 turned away from them by extending at an angle from the channel area CHANNEL. This makes it possible to derive electrical signals from a single biological cell with a large number of sensor transistors.

4 shows a further preferred embodiment of a sensor device according to the invention. The embodiment is similar to that of 3 , a backgate electrode being arranged on a rear mesa side surface of the channel region CHANNEL of the FINFET. The backgate electrode 28 , which in a known manner by a gate dielectric 30 is spaced from the channel region CHANNEL of the FINFET, is thus arranged on the mesa side surface of the FINFET that corresponds to the sensor surface 16 opposite. In the case of not too thin body areas (channel areas) and in particular for channel thicknesses (distance between vertical mesa side surfaces of the FINFET in the channel area) of more than 50 to 100 nm, the resulting component can be 5 shown equivalent circuit diagram are modeled. With the backgate electrode 28 A further transistor channel can thus be formed in the channel region CHANNEL, in which the “sensor transistor” is also designed for the detection of the electrical signals of the cell to be examined. This additional transistor can be used to compensate for parameter variations of the "sensor transistor", particularly in sensor array applications. With suitable control, process-related parameter variations of different FINFETs can be compensated for by the additional transistor.

The calibration circuit or the calibration method is based on the fact that, in a calibration phase, charge is applied to a gate node of the backgate electrode 28 is stored, the size of the stored amount of charge compensating for the variation of the "sensor transistors" within the sensor array. Preferably in the calibration phase, which in 5 (a) is shown, a fixed voltage is applied to the electrolyte bath with respect to the source contact of the FINFET. The switch S_cal, which can be implemented, for example, by a further transistor, is closed and a current I_cal is impressed into the given circuit configuration. The potential on the backgate electrode 28 under this constraint, the transistors T1 (sensor transistor) and T2 (calibration transistor) carry the current I_cal regardless of the deviations of their characteristic parameters from the mean value of these parameters. The switch S_sel is symbolically inserted in the line between the circuit comprising the transistors T1 and T2 and the current source and symbolizes one or more selection switches which are required for the position selection when the configuration from T1 and T2 is operated in a sensor array. The switch S_sel is also usually implemented using field effect transistors.

During the measurement phase, the switch S_cal is open and on the backgate electrode 28 Exactly the amount of electrical charge remains, which causes the potential that was established at this node during the calibration phase. Variations in the control voltage caused by the cell under test 24 are effected, transistor T1 now acts as current modulation of the total current of T1 and T2 around the value I_cal. Unless the cell 24 emits no electrical signal, the parallel connection from the transistors T1 and T2 carries exactly the current I_cal. In read mode, the circuit can work against a load resistor, for example, so that a voltage signal can be tapped (cf. 5 (b) ) or a current source so that a differential current can be tapped at the output of the circuit (cf. 5 (c) ).

In the event that the channel or body area is made very thin, this applies in 5 shown equivalent circuit diagram of two (independent) transistors no longer. In this case, a weighted sum of the cell signal and the voltage of the back gate electrode acts 28 on the channel area of a transistor. The possibility of calibration via charge storage on the backgate electrode 28 however, also exists in this case.

A particularly preferred property of sensor devices according to the invention is to enable the biological cells to be examined to be re-linked to a neural network. In order to be able to record the entire network properties, the sensor device must be able to operate several sensor transistors with coupling elements and sensor areas in parallel. From a certain minimum number of "sensor cages", it is therefore necessary to change the number of access lines for the sensor transistors via appropriate circuits, as in for example, multiplexers to reduce. Sensor devices which have a monolithically integrated evaluation and control unit in the immediate vicinity of the sensor transistors are therefore particularly preferably used. Such an embodiment is in a schematic sectional view in 7 shown. This embodiment of a sensor device according to the invention has a cell array with vertical transistors, which is denoted overall by 32. At the in 7 In the embodiment shown, the section plane runs perpendicular to the substrate surface plane through two cell spaces 18 therethrough. The vertical sensor transistors used will be related later 8 (c) described in more detail.

To the cell field area 32 an evaluation and control electronics area closes 34 which is preferably carried out in conventional, planar CMOS technology. The evaluation and control electronics area 34 is opposite the cell field area 32 through trench insulation 36 isolated. A large number of horizontal bulk transistors with a plurality of wiring levels above them are provided in the evaluation and control electronics area, with which switches and amplifiers in particular can be integrated in the immediate vicinity of the sensor cages. For example, in 7 Embodiment shown the drain region DRAIN of a sensor transistor via an electrically conductive bridge element 38 with a gain or impedance conversion transistor 40 connected. The impedance conversion or amplification transistor is a conventional, planar CMOS transistor.

In 8th Further preferred embodiments of sensor devices according to the invention are shown in schematic sectional views. The section plane runs perpendicular to the substrate surface plane through source and drain regions SOURCE, DRAIN of the sensor transistors. At the in 8 (a) embodiment shown is the cell 24 into a cell room 18 placed, which is designed as a trench. The source area SOURCE represents the bottom of the trench and forms the contact surface 26 , Optionally, a tapered element 42 be provided which tapers the trench on both sides or partially on one side in order to fix the biological cell in place 24 to improve. In order to produce this additional trench taper, the trench must first be filled with a second auxiliary layer, for example a nitride layer. The tapered layer, which is preferably an oxide layer, is then structured. The nitride layer is then selectively removed from the oxide layer, for example by means of wet etching in hot phosphoric acid. At the in 8 (a) In the embodiment shown, the source and drain regions SOURCE, DRAIN are doped by ion implantation. The trench base, which of the contact surface 26 is smaller than the trench height in the substrate normal direction in this embodiment.

At the in 8 (b) In the embodiment shown, both the source and drain regions SOURCE, DRAIN lie in an upper semiconductor layer. In this arrangement, the vertical, ie parallel to the substrate normal direction trench side surfaces represent the sensor surfaces 16 However, the contact surface also has an effect 26 as a sensor surface 16 , At the in 8 (c) In the embodiment shown, the drain region DRAIN is formed in an upper semiconductor layer. The two trench side surfaces form mutually independent sensor transistors which have a common source region, which is shown as a buried semiconductor layer. The investment area 26 is arranged in the common source region layer. At the in 8 (d) shown embodiment, a pyramid-shaped trench etching is used, thereby creating a pyramid-shaped cell space 18 with tapered element 42 is created. 8 (e) and (F) show possible configurations of the source contacts 20 and drain contacts 22 , At the in 8 (e) The embodiment shown is a common source contact (not shown) and individual drain contacts 22 used which opposite the cell room 18 (the trench) are set back. At the in 8 (f) The embodiment shown is the contacting of the common source area SOURCE by a source contact 22 shown. In addition, this embodiment also has a body contact 44 which can be used to avoid the floating body effect.

All Exemplary embodiments described above are freely combinable with each other, so that elements that only in the context of an embodiment have also been used in other embodiments can.

DRAIN

drain region

CHANNEL

channel area

SOURCE

source region

SUB

Semiconductor substrate

10

Semiconductor layer for source area

12

Semiconductor layer for channel area

14

Semiconductor layer for drain area

16

Sensor area of the coupler

18

cell space

20

source contact

22

drain contact

24

biological cell

26

On- or contact surface

28

Back gate electrode

30

gate dielectric for backgate electrode

32

Cell array region

34

evaluation and control electronics area

36

trench isolation

38

bridge element

40

amplifying transistor

42

rejuvenation element

44

Body Contact

Claims (15)

Sensor device for the detection of electrical signals of at least one biological cell ( 24 ) comprising - a flat semiconductor substrate (SUB) extending along a plane of the substrate surface with a contact surface ( 26 ) to which the biological cell ( 24 ) can be created; - At least one field effect transistor formed on the semiconductor substrate (SUB), which has a source, a drain and an intermediate channel region (SOURCE, DRAIN, KANAL); - at least one coupling element adjacent to the channel area (CHANNEL) with a sensor surface ( 16 ), with a normal direction of the sensor surface ( 16 ) is inclined at least in some areas with respect to a normal direction of the substrate surface plane, so that the sensor surface ( 16 ) from the contact surface ( 26 ) protrudes; the coupling element being designed such that the electrical signals to the sensor surface ( 16 ) adjacent biological cell ( 24 ) influence the electrical conductivity of the channel area (CHANNEL) via a capacitive coupling. Sensor device according to claim 1, wherein the coupling element consists of at least one dielectric layer. Sensor device according to claim 2, wherein the coupling element has a layer thickness in the range from 1 nm to 100 nm. Sensor device according to one of the preceding, wherein the contact surface ( 26 ) is substantially parallel to the plane of the substrate surface. Sensor device according to one of the preceding claims, wherein the normal direction of the sensor surface ( 16 ) is inclined relative to the normal direction of the substrate surface plane by an angle in the range from 30 to 130 degrees, preferably about 90 degrees. Sensor device according to one of the preceding claims, wherein the sensor device has a plurality of field-effect transistors, each with one of the coupling elements, and the coupling elements are arranged in such a way that the sensor surfaces ( 16 ) of the coupling elements at least partially a cell space ( 18 ) in directions parallel to the substrate surface plane. Sensor device according to claim 6, wherein the normal directions of the sensor surfaces ( 16 ) almost in a central area of the cell room ( 18 ) to cut. Sensor device according to one of the preceding Expectations, wherein the semiconductor substrate is an SOI substrate and the field effect transistor is a FINFET. Sensor device according to claim 8, wherein the source and the drain region (SOURCE, DRAIN) of the FINFET in a direction parallel to the normal direction of the sensor surface ( 16 ) from the sensor surface ( 16 ) stretch away. Sensor device according to one of claims 8 or 9, wherein the channel region (CHANNEL) has two opposite interfaces and on one of the interfaces the coupling element and on the other of the interfaces a back gate electrode ( 28 ) is arranged. Sensor device according to one of claims 1 to 7, wherein the field effect transistor is a vertical transistor, in which the source, the channel and the drain region (SOURCE, KANAL, DRAIN) in semiconductor layers ( 10 . 12 . 14 ) are formed which adjoin one another in the normal direction of the substrate surface plane. Sensor device according to claim 11 with claim 6 or 7, wherein all of the source regions (SOURCE) of the field effect transistors in the same semiconductor layer ( 10 ) are designed as a common source area. Sensor device according to one of the claims che 11 or 12, whereby at the channel area (CHANNEL) of the FINFET an ohmic body contact ( 44 ) is trained. Sensor device according to one of the preceding claims, wherein on the semiconductor substrate (SUB) additional evaluation electronics ( 34 ) is designed for the electrical signal to be detected. Method for the detection of electrical signals of at least one biological cell ( 24 ) comprising the steps: - providing a sensor device according to one of the preceding claims; - suspending the biological cells ( 24 ) in a solution; - Applying the solution to the sensor device so that one of the biological cells ( 24 ) to the contact surface ( 26 ) and the sensor surface ( 16 ) adjoins; and - detecting the electrical signals of the cell ( 24 ) about their influence on the electrical conductivity of the channel area (KANAL) by measuring the electrical conductivity of the channel area (KANAL).