EP4260055A1

EP4260055A1 - Ferroelectric biochip

Info

Publication number: EP4260055A1
Application number: EP21839004.5A
Authority: EP
Inventors: Maximilian Becker; Claus Burkhardt
Original assignee: NMI Naturwissenschaftliches und Medizinisches Institut
Current assignee: NMI Naturwissenschaftliches und Medizinisches Institut
Priority date: 2020-12-09
Filing date: 2021-12-07
Publication date: 2023-10-18
Also published as: WO2022122756A1; DE102020132756B3

Abstract

The invention relates to a biochip (13) having at least one coupling arrangement (21) for electrically stimulating biological material (24) or for conducting electrical measurements on the biological material (24), the biochip (13) comprising: a carrier structure (23) arranged on and/or in the coupling arrangement (21); and a layer (17) whose one layer surface (19) is arranged on the coupling arrangement (21) and whose opposite layer surface forms a coupling surface (15) for stimulating the biological material (24) and/or for conducting measurements on the biological material (24). To obtain an improved stimulation efficiency of the biochip (13), the layer (17) is proposed to have ferroelectric properties.

Description

Ferroelectric biochip

The invention relates to a biochip with at least one coupling arrangement for electrical stimulation of biological material or for electrical measurements on the biological material.

[0002] Patent EP 1 478 737 B1 discloses a biochip which is set up for the capacitive stimulation and/or detection of biological tissue. In order to improve the coupling efficiency of the purely capacitive coupling provided in this biochip, it is proposed to use a dielectric with the highest possible relative permittivity e _r , namely TiO ₂ .

[0003] There are also semiconductor components that have regions made of a ferroelectric material. Corresponding manufacturing processes are also known. An example of such a semiconductor device is the ferroelectric field effect transistor (FeFET). Such semiconductor components are used, for example, to produce non-volatile semiconductor memories.

[0004] Despite the progress made in the field of biochips with purely capacitive coupling, their stimulation efficiency is generally not sufficient to implement, for example, microelectrode arrays (MEAs) or electrically active implants with a high spatial density of electrodes. The stimulation efficiency determines the minimum area of each electrode for which effective electrical stimulation of the biological material, which may contain nerve cells, is practically possible.

[0005] One object of the present invention can therefore be seen as providing a biochip with improved stimulation efficiency, with which, for example, microelectrode arrays or electrically active implants with a high density of electrodes can be implemented, electrochemical degradation processes on the electrodes also being suppressed should be. According to one embodiment of the invention, a biochip with at least one coupling arrangement for electrically stimulating biological material or for electrically measuring the biological material is proposed, the biochip having: a carrier structure on and/or in the at least one coupling arrangement is arranged; and a layer, one layer surface of which is arranged on the coupling arrangement and the opposite layer surface of which forms a coupling surface for electrical stimulation of the biological material and/or for electrical measurements on the biological material; which is characterized in that the layer has ferroelectric properties. In other words, the layer of such a biochip has a non-linear and/or hysteretic relationship between an electric field E within the layer and an electric displacement density D, which is caused by the presence of a ferroelectric polarization is caused within the layer. This non-linear and hysteretic behavior of this connection DE) is also referred to as ferroelectric behavior. In this sense, materials with ferroelectric properties also include antiferroelectric materials, ferrielectric materials, multiferroic materials or relaxor ferroelectric materials. The layer can therefore have ferroelectric, antiferroelectric, ferrielectric and/or relaxor ferroelectric materials or also be formed from one of these materials.

The inventors have recognized that the layer with the ferroelectric properties makes an additional contribution to the electrical stimulation of the biological material, which results from the ferroelectric polarization in the layer. This additional contribution occurs in addition to a contribution caused by the known purely capacitive stimulation. In many cases, the additional contribution can even be higher than the contribution of the purely capacitive coupling.

The layer can have, for example, hafnium oxide (HfO ₂ ) with ferroelectric properties or be formed therefrom. Such ferroelectric hafnium oxide is compatible with known semiconductor processes, such as CMOS technology, so that simple production of the biochip is possible. A further improvement in the properties of the biochip and/or a simplification of the manufacturing process can be achieved by doping hafnium oxide, for example if the hafnium oxide having ferroelectric properties is doped with silicon. Alternatively, the hafnium oxide can also be doped with one of the following materials: aluminum (Al), germanium (Ge), yttrium (Y), gadolinium (Gd), lanthanum (La) or strontium (Sr).

For example, the hafnium oxide having ferroelectric properties can also be doped with zirconium (Zr). This means that the layer comprises or is formed from a material of the class Hf _1.x ZrxO ₂ .

The material is preferably Hf _{0 5} Zro. ₅ 0 ₂ . This material has a high remanent polarization P _r compared to other dopings with zirconium, as a result of which the additional contribution to electrical stimulation of biological material described above is particularly pronounced.

It can also be provided that the layer has zirconium oxide (ZrO ₂ ) with ferroelectric properties or is formed from such.

Furthermore, it can be provided that the layer has aluminum scandium nitride with ferroelectric properties or is formed from aluminum scandium nitride. Aluminum scandium nitride is to be understood as class A .xSCxN materials. On the one hand, this class of materials is compatible with common manufacturing processes that are also used to produce CMOS circuits, and on the other hand it has a particularly high remanent polarization of over 100 pC/cm ² .

It can also be provided that the layer has a ferroelectric perovskite and/or ferroelectric polymer or is formed from such.

The applicability of the technical teaching described here is not limited to materials with purely ferroelectric properties. It can also be provided that the layer has multiferroic properties. So the layer can be a multiferroic Have material or be formed from such. The multiferroic material can be bismuth ferrite (BiFeO ₃ ), for example.

It goes without saying that any material systems mentioned in the prior art, such as barium titanate (BaTiO ₃ ) or PZT (lead zirconate titanate, PbCZrxTh. Os) do not fundamentally have ferroelectric properties. This is abbreviated and inaccurately simplified in some publications. Material systems such as barium titanate and PZT can be designed in such a way that they have ferroelectric properties. However, this is by no means always the case (see for example DJ McClure and JR Crowe; "Characterization of amorphous barium titanate films prepared by rf sputtering", Journal of Vacuum Science and Technology 16, 311 1979 for barium titanate without ferroelectric properties; and Hu et al. , "EXAFS study of PZT ferroelectric thin films of different crystallinities", Journal of Physics: Conference Series 430 (2013), 15th International Conference on X-ray Absorption Fine Structure (XAFS15), 2013 for PZT without ferroelectric properties). In the case of layers in particular, such as thin dielectric layers with amorphous (possibly metallic) materials, it is not necessarily the case that these layers are formed in such a way that they necessarily have ferroelectric properties. If a material that can have ferroelectric properties as a “bulk” material (e.g. PZT or barium titanate) is produced as a layer or thin film, it is not necessarily the case that the produced layer or thin film also has ferroelectric properties. However, it is also possible to provide a layer or thin layer in such a way that the layer or thin layer has ferroelectric properties. The ferroelectric properties can be verified by measurement. Optionally, the layer having ferroelectric properties is not made of barium titanate and/or the layer having ferroelectric properties is not made of PZT. Optionally, the ferroelectric property layer does not include barium titanate and/or the ferroelectric property layer does not include PZT.

In particular in the case of electrical stimulation of the biological material, the thickness of the layer can be greater than 200 nm and preferably in the range between 200 nm and 1500 nm, in particular between 500 nm and 1500 nm or between 200 nm and 800 nm . Unlike known biochips, which are exclusively have capacitive coupling and are therefore dependent on the smallest possible layer thickness to achieve the greatest possible coupling capacitance, the layer thickness can be selected to be relatively high in the biochip described here and good stimulation efficiency can nevertheless be achieved. The high layer thickness ensures good electrical insulation between the biochip and the biological material, ie low leakage currents. Furthermore, a biochip with a comparatively large layer thickness is more stable over the long term.

[0018] If the carrier structure has a substrate made of a semiconductor material or is formed from such a material, the biochip can be easily manufactured using manufacturing processes known from semiconductor technology. It can also be provided that the carrier structure has at least one of the following materials or is formed from it: polyimide, epoxy resin, parylene. These materials are biocompatible and therefore particularly suitable for producing the biochip. However, other biocompatible and/or dielectric materials can also be used.

In particular, if the biochip is (also) set up for electrical measurements on the biological material, the layer can form part of a ferroelectric field effect transistor (FeFET) of the biochip. In this case, the FeFET can form part of an input stage of a corresponding measuring circuit. It can also be provided that the FeFET belongs to an artificial neuron. This can, for example, be coupled to a biological neuron present in the biological material.

The procedure can be that the coupling surface has a gate electrode of the ferroelectric field effect transistor or forms one. In this case, the layer can correspond to a ferroelectrically acting insulating layer of the FeFET. Alternatively, a multi-layer construction of the gate area of the FeFET is also conceivable, with the gate area comprising the ferroelectric layer and a further layer, which can be an insulating layer.

The biochip can be designed in such a way that it has memresistive or memristive properties. In particular, the layer can be formed in such a way that a ferroelectric tunnel contact is formed. The biochip can be designed as a memristor be. In particular when the layer with ferroelectric properties is particularly thin, the memristive properties can occur in particular in that a so-called ferroelectric tunnel contact is formed. Memristive properties can then occur due to the ferroelectric properties of the layer between the electrodes: for example, because a ferroelectric tunnel contact is formed. The biochip can therefore be set up in such a way that a ferroelectric tunnel contact is formed. A thickness of the layer with ferroelectric properties is, for example, less than or equal to 10 nm, in particular less than or equal to 7.5 nm, in particular less than or equal to 5 nm. A biochip designed as a memristor can be used as an artificial synapse.

The biochip can be set up to be used as part of an arrangement which couples at least one artificial neural network to a biological neural network. The biochip can be set up to couple an artificial neuron to a biological neuron.

[0023] The biochip can have a plurality of mutually separate coupling surfaces of different coupling arrangements of the biochip. A coupling surface can be assigned to each coupling arrangement. A multi-channel biochip can be provided in this way. The coupling surfaces can be arranged, for example, along a line or in a grid. Such a biochip can have several separate layers with ferroelectric properties.

According to a further embodiment, a method for producing a biochip with at least one coupling arrangement for electrical stimulation of biological material or for electrical measurements on the biological material is proposed, the method comprising: arranging a carrier structure on and/or in the coupling arrangement; and producing a layer in such a way that one layer surface is arranged on the coupling arrangement and the opposite layer surface forms a coupling surface for electrical stimulation of the biological material and/or for electrical measurements on the biological material, the method being characterized in that the layer is produced as a layer with ferroelectric properties. With such a method, the above in connection with Realize the advantages described for the biochip. The features of the biochip can be provided in the method in a corresponding manner.

A further embodiment comprises an electrically active implant with the biochip described here.

The implant can be, for example, a retina implant, a cochlear implant, an implant for deep brain stimulation and/or an implant for providing a brain-machine interface. In general, implants for stimulating nerve cells and/or for measurements on nerve cells can be provided, such implants having at least one biochip described here.

As yet another embodiment, an in vitro arrangement with a biochip described here and an electrolyte receptacle, i.e. an arrangement for receiving an electrolyte, is disclosed, the electrolyte receptacle for arranging the electrolyte on at least one coupling surface of the biochip for electrical stimulation of the biological material and/or for electrical measurement of the biological material.

[0028] An in vitro arrangement can hereby be provided which has at least one biological neuron and at least one artificial neuron. The biological neuron can be formed, for example, by a nerve cell of the biological material located in the electrolyte. The artificial neuron can comprise an FeFET, which can be part of the biochip and can be implemented as described above.

The implant or the in-vitro arrangement can have a semiconductor chip separate from the biochip, a connection of the semiconductor chip being electrically connected via a conductor track to an electrode layer adjoining the layer surface, and the conductor track and the biochip being arranged on a common substrate .

Other features and advantages emerge from the following description. Here show: [0031] FIG. 1 shows a schematic representation of an arrangement with a biochip;

[0032] FIG. 2 shows the basic structure of the biochip from FIG. 1;

[0033] FIG. 3 shows a stimulation device of the biochip from FIG. 1;

[0034] FIG. 4 shows how the stimulation device from FIG. 3 works;

[0035] FIG. 5 shows a graphic representation of the time profile of a possible voltage signal U(t) for the stimulation device from FIG. 3;

6 shows a graphical representation of the non-linear and hysteretic relationship D(E) between the electric field and electric displacement density of a ferroelectric layer of the biochip;

FIG. 7 shows a representation similar to FIG. 6, but for the case of an antiferroelectric layer;

[0038] FIG. 8 shows a measuring device of the biochip from FIG. 1;

[0039] FIG. 9 shows a further measuring device of the biochip from FIG. 1; and

[0040] FIG. 10 a biochip arrangement with a measuring device.

1 shows an arrangement 11 with a biochip 13. In the example shown, the arrangement has only one biochip 13. FIG. However, it is also possible to provide several biochips, preferably arranged next to one another in a grid, on or in the arrangement 11 . The arrangement 11 has a plurality of coupling surfaces 15 . Depending on the precise configuration, at least one of these coupling surfaces 15 can be set up for the electrical stimulation of biological material, in particular a nerve cell located therein. In addition, at least one of these coupling surfaces 15 for electrical measurements on the biological material, in particular on a nerve cell located therein. The biochip 13 can therefore have a plurality of coupling surfaces 15 , with each coupling surface 15 being assigned to exactly one coupling arrangement 21 of the biochip 13 . The coupling surfaces 15 and/or the coupling arrangements 21 can be arranged along a line or, as shown in FIG. 1, in a grid.

The arrangement 11 can, for example, be an electrically active implant such as a neuroprosthesis. The arrangement can be a cochlear or retina implant. Furthermore, the arrangement 11 can be an implant for deep brain stimulation for the treatment of neurological symptoms such as Parkinson's disease. In such neuroprostheses 11 it can be provided that all coupling surfaces 15 of the at least one biochip 13 are set up for stimulating nerve cells.

Alternatively, the arrangement can be an in vitro arrangement comprising a microelectrode array (MEA) whose microelectrodes correspond to the coupling surfaces 15 . Such an in-vitro arrangement can have at least one biological neuron, which is formed by a nerve cell located in the biological tissue, and/or at least one artificial neuron, which belongs to the biochip 13. Such an arrangement can be used, for example, to study neuronal networks or nerve tissue. The biological material can be in an electrolyte, which can be brought into contact with the biochip 13, in particular its coupling surfaces 15, via a suitable electrolyte receptacle of the arrangement 11, such as an electrolyte container 16, for example.

Known arrangements 11 (for example neuroprostheses and not purely capacitive MEAs) have electrically conductive (metallic) electrodes. Such electrodes enable a high charge transfer per unit area (stimulation efficiency) of electrical charge between the biological material and the biochip 11. The high stimulation efficiency allows the area of the individual electrodes to be kept small, which in turn allows a high area density of electrodes. However, it should be noted that a threshold value required for the successful stimulation of nerve cells is the the charge density to be transmitted by the nerve tissue depends on the size of the electrodes or the coupling surfaces 15 .

A disadvantage of conductive electrodes is that irreversible electrochemical processes in the form of a Faraday current can occur at the electrode/electrolyte interface, which can lead to degradation of the electrode or to cell damage in the biological material. The Faraday currents at the electrodes can be avoided, at least to a large extent, if a biochip with purely capacitive coupling is used. Such a biochip usually has conductive areas (for example p-doped areas in silicon) which are covered with a dielectric layer which suppresses the Faraday currents. However, the stimulation efficiency of these biochips is relatively low and is determined in particular by the relative dielectric constant of the dielectric layer and its thickness. Attempts are therefore being made to use materials with the highest possible relative dielectric constant and to keep the thickness of the layer small. However, a small layer thickness can lead to increased leakage currents, which counteracts the goal of reducing irreversible processes.

In the arrangement 11 described here, the biochip 13 has at least one layer 17 with ferroelectric properties, which is delimited by the coupling surface 15 .

The basic structure of the biochip 13 is shown in detail in FIG. It can be seen that a layer surface 19 of the layer 17 opposite the coupling surface 15 adjoins a carrier structure 23 of the coupling arrangement 21 of the biochip 13 . The coupling surface 15 and the layer surface 19 form opposite layer surfaces of the layer with ferroelectric properties 17. As far as the precise configuration of the carrier structure 23 is concerned, there is a great deal of design freedom. For example, as shown in FIG. 2, the carrier structure 23 can consist of one layer. This layer can have, for example, conductive areas which consist of a semiconductor material which is p-doped or n-doped in order to achieve a desired electrical conductivity. However, the carrier structure 23 can also have metallic conductor tracks which are arranged in or on the carrier structure 23 . Furthermore, the support structure 23 consist of a flexible and biocompatible material such as polyimide.

The coupling surface 15 of the layer 17 delimits the biochip towards an area 25 in which the biological material is located. The biological material can be in an electrolyte 24 . The biochip 13 and the area 25 with the biological material are components of the biochip arrangement 11. The biological material can have at least one nerve cell which can form a biological neuron.

In the example shown, the coupling surface 15 forms an electrode with a substantially smooth surface. In an example not shown, the coupling surface 15 is machined to have an increased surface area. As a result, an actively effective surface can be enlarged in order to further increase the charge transfer. The coupling surface 15 can be circular overall with a diameter of 5 μm to 200 μm, preferably 5 μm to 100 μm, in particular 30 μm. The coupling surface 15 can also have another shape, for example a rectangle, square or other polygon shape (preferably with the same surface area as the circular coupling surface). If several coupling surfaces 15 are provided per biochip 13, they can have a minimum distance of 5 μm to 500 μm from one another.

3 shows a cross section of a biochip that is set up for electrical stimulation of the biological material. The carrier structure 23 has two layers here, namely an electrically conductive electrode layer 27 directly adjoining the layer surface 19, which can be formed from metal or doped semiconductor material, and a substrate layer 29 adjoining the electrode layer 27. The electrode layer 27 is therefore located between the layer 17 and the substrate layer 29. The electrode layer 27 is arranged on the layer 17 so that it interacts with the material of the layer 17 having ferroelectric properties, in particular when a time-varying electrical voltage U(t) is applied to the electrode layer 27. To apply the voltage, a suitable contact can be provided, as shown schematically in FIG. 3, via which a voltage source 31 can be connected to the electrode layer 27. Furthermore, the biochip arrangement 11 a Have counter-electrode 33 with which the biological material and/or the electrolyte 24 can be contacted.

If the biological material (e.g. a nerve cell) in the electrolyte 24 is to be stimulated, an electrical voltage U(t) is applied to the electrode layer 27 acting as an electrode and to the counter-electrode. In contrast to devices with exclusively electrically conductive electrodes, the electrolyte 24 is electrically insulated from the electrode layer 27, so that no direct current can flow, which leads to an electrochemical charge transport.

As shown in FIG. 4, due to the change over time in the voltage U(t) generated by the voltage source 31, there is still a charge shift or a current flow in the electrolyte 24, which is used for electrical stimulation of the biological material located therein . As with purely capacitive stimulation, the current flow in the electrolyte corresponds to a Maxwellian displacement current. dD

The displacement current density is given as the time derivative J _D = For a

Material with ferroelectric properties is O (£), where s ₀ denotes the permittivity of the vacuum.

In the case of a ferroelectric layer, the displacement current density J _D dd (FE

= EQE^F + ^P (F) in the electrolyte 24 consists of two parts. A proportion is caused by a temporal change in the dielectric polarization P ^{(D 1 )} , which in turn is caused by the electric field E(t) within the layer 17 corresponding to the voltage U t ). This purely dielectric displacement current density is proportional to the dielectric constant e _r of layer 17 and forms the only contribution in the purely capacitive stimulation of biological material. In the case of a layer with ferroelectric properties, however, there is an additional component which depends on the time ■■ (FE

Change in the ferroelectric polarization P is caused, which in turn is also of the voltage U (t) corresponding electric field E (t) within the Layer 17 is caused. For a layer 17 of thickness d (see Fig. 2) defined by . . U(t) is limited to the two layer surfaces 15, 19, the relationship E(t)=

Fig. 5 shows an example of the time profile of a bipolar voltage signal in the form of a sinusoidal oscillation with a frequency f of 100 Hz and an amplitude U _max = - U _min , a value of the voltage signal during its time profile a maximum value U _max and a assumes a minimum value U _mjn , which corresponds to a maximum value E _max and a minimum value E _min of the corresponding electric field strength E within the layer 17 . Such voltage signals can be used, for example, for the electrical stimulation of electrogenic cells, but they can also be used to measure the nonlinear and hysteretic relationship D(E) of materials with ferroelectric properties.

6 shows the typical course of the non-linear and hysteretic relationship D(ε) for a ferroelectric material. The _{ferroelectric} polarization can be reversed by applying a field strength E which is above the material-specific coercive field strength Ec. After switching off the electric field, the remanent FE remains)

Polarization P _r = P (0) obtained, which can assume the two states +P _r and -P _r depending on the history. Ferroelectric memories are based on these two stable states.

7 shows the typical course of the nonlinear and hysteretic relationship D(E) for an antiferroelectric material. The ferroelectric hysteresis occurs here only above a critical field strength EC _r n _D. After switching off the electric field, the electric polarization disappears again.

The stimulation efficiency can be quantified from the knowledge of the non-linear and hysteretic relationship D(E) of a material with ferroelectric properties shown in FIG. 6 and FIG. If the stimulation efficiency is transferred with a charge density p _stim quantified (caused by a voltage signal U(t) such as that in Fig. 5 and measured e.g. in pC/cm ² ), the following general relationship results p “sttm = D(E max )' - D(E mi .n )'.

In the case of polarity reversal of a ferroelectric material which, for example, has the hysteresis curve D(E) shown in FIG. 6, it follows from the general relationship for the stimulation efficiency

p _stim = C _s U _max - U _min ) + 2P

Here, C _s corresponds to the capacitance per unit area (measured e.g. in pF/cm ² ) of a plate capacitor formed by the two layer surfaces 15, 19 and P _r corresponds to the amount of remanent polarization of the ferroelectric material of layer 17.

The proportion 2P _r of the transmitted charge density p _stim that originates from the ferroelectric properties results from the fact that the voltage U(t) generated by the voltage source 31 causes a polarity reversal of the remanent polarization P _r in the layer 17 . In order to control the biochip 13, the amplitude of the voltage signal U(t) can be selected such that the corresponding field strength E in the layer 17 is at least as great in absolute terms as the coercive field strength E _c .

The layer 17 can consist of hafnium oxide (HfO ₂ ) with ferroelectric properties or doped hafnium oxide with ferroelectric properties or at least contain the same.

For example, the layer 17 can consist of zirconium-doped hafnium oxide (Hf _1.x Zr _x O ₂ ) or at least contain the same. For x = 0.5, the result is Hf _0.5 Zro. ₅ 0 ₂ , with which the layer 17 can be produced with a thickness of d=9.5 nm. Such a layer 17 has the following properties at U _min = 0V and U _max = IV: e _r = 40, C _s = 3.7 pF/ _cm ² , E _c = 100 or 1000 ^kV /cm, P _r = 16 pC/ _cm ² ,

It can be seen that the proportion 2P _r in this example is already 32 |iC/ _cm ² , ie it far exceeds the proportion achieved by purely capacitive stimulation.

The layer 17 can also consist of aluminum scandium nitride (A.xSCxN) with ferroelectric properties or at least contain the same. This material is compatible with semiconductor manufacturing processes that are also used to make CMOS circuits. Because of its relatively high remanent polarization of over 100 nC/cm, transferrable charge densities of more than 200 p.C/cm result.

The material from which the layer 17 consists or which the layer 17 comprises can also be an antiferroelectric. The material can be, for example, hafnium oxide (HfO ₂ ) doped with silicon (Si). Antiferroelectrics have a curve D(E') shown qualitatively in FIG.

Finally, the material which has the layer 17 or from which the layer 17 consists, can also have multiferroic properties, ie it can be both ferroelectric and ferromagnetic. For example, this can be bismuth ferrite (BiFeO ₃ ).

Since the stimulation efficiency is largely determined by the ferroelectric properties, the thickness d of the layer 17 can be selected to be comparatively large. Instead of the above value of d=9.5 nm, significantly larger values are also possible, e.g. B. more than 200 nm, 200 nm to 1500 nm, 200 nm to 800 nm or 500 nm to 1500 nm.

However, another effect can also be used. If the layer with ferroelectric properties is particularly thin, the memristive properties can occur in particular in that a so-called ferroelectric tunnel contact is is formed. The biochip can therefore be set up in such a way that a ferroelectric tunnel contact is formed. A thickness of the layer with ferroelectric properties is, for example, less than or equal to 100 nm, in particular less than or equal to 50 nm, in particular less than or equal to 20 nm, in particular less than or equal to 10 nm, in particular less than or equal to 7.5 nm, in particular less than or equal to 5 nm. in particular less than or equal to 4 nm, in particular less than or equal to 2.5 nm.

It goes without saying that the ferroelectric layer (17) can be arranged not only exclusively in a planar (2D) plate capacitor design, but also in other arrangements such as, for example, in a 3D capacitor structure such as a so-called deep trench capacitor or a deep trench capacitor.

FIG. 8 shows a section of a biochip 13 that is set up to carry out electrical measurements on the biological material located in the electrolyte 24 . The carrier structure 23 has a substrate layer 29 made of semiconductor material. A drain zone 35 and a source zone 37 are arranged in the substrate layer 29, preferably in a region which adjoins the layer 17. FIG. The two zones 35, 37 are arranged on the layer 17 at a distance from one another. The two zones 35 and 37 can be formed in that the semiconductor material there is oppositely doped with respect to a doping of the substrate layer 17 outside of the zones 35, 37. For example, the two zones 35, 37 can be n-doped and the rest of the substrate layer 29 can be p-doped (or vice versa).

The substrate layer 29, the drain region 35 and the layer 17 form a ferroelectric field effect transistor (FeFET). In this case, the coupling surface 15 of the layer 17 corresponds to a gate contact of the FeFET 39 which is electrically connected to the electrolyte 24 . The layer 17 forms an insulating layer of the FeFET 39, which is comparable to a gate oxide of a classic MOSFET. The drain zone 35, the source zone 37 and the substrate layer 29 are electrically connected to corresponding terminals U _D , U _S and U _B of the FeFET 39 which each have a drain contact, a source contact and a bulk contact of the FeFETs 39 form. The electrolyte 24 can be brought to a reference potential, for example a ground potential GND, by means of a reference electrode. It can be provided here that the bulk connection U _B of the FeFET is at the same electrical potential as the electrolyte 24 and/or that the reference electrode is integrated into the substrate layer 29.

The FeFET 39 can form part of an input stage of the biochip arrangement 11, which is designed for electrical measurements on biological material, in particular on nerve cells contained therein. Furthermore, the FeFET 39 can also be a part of an artificial neuron. A total of 11 examinations of nerve tissue can be carried out using the biochip arrangement. The biochip arrangement 11 can be used for neuromorphic computing. In addition, the biochip arrangement 11 can be a brain-machine interface (brain-machine interface), in particular a brain-computer interface (brain-computer interface).

A measuring device of the biochip 11 can also be constructed as shown in FIG. Unlike the measuring device shown in FIG. 7, the carrier structure 23 has an electrical connection arrangement (“metal stack” 41). An insulating layer 47 is assigned to the drain zone 35 and the source zone 37, so that a MOSFET structure 42 comprising the two zones 35, 37 and the insulating layer 47 results. The insulating layer 47 can be arranged on the substrate layer 29 and/or consist of polysilicon. The connection arrangement 41 is arranged between the insulating layer 47 and the ferroelectric layer 17 in such a way that it electrically connects them to one another. In the example shown, six metal layers (“metal layers”) are provided in the semiconductor technology used. Correspondingly, the connecting arrangement 41 has six conductors 43 arranged in different metal layers, conductors of adjacent metal layers being connected to one another via plated-through holes (“vias” 45). Deviating from this, the number of metal layers can be varied as required.

Another example of a measuring device is the arrangement 11 shown in FIG. 10, which is particularly suitable for carrying out in-vivo measurements. Similar to the measuring device from FIG. 9, the measuring device has a MOSFET. However, this belongs to a semiconductor chip 49 that is separate from the biochip 13 . A connection of the semiconductor chip 49 is electrically connected to the electrode layer 27 via a conductor track 43 . The connection can be a gate connection of the MOSFET. In the example shown, the conductor track 43 is on the substrate 29 of the biochip 13 appropriate. The substrate 29 may be formed from a biocompatible dielectric material such as a polyimide. In this case, the conductor track 43 and the substrate 29 can form a flexible printed circuit board (“flex PCB”). The conductor track 43 can consist of gold, for example, and/or be electrically insulated on a side facing away from the substrate 29, for example by means of a parylene layer.

In the measuring device, the elements 43, 27 and 17 can be present multiple times and can be arranged next to one another, for example in a direction which runs at least essentially orthogonally to the plane of the drawing in FIG. This results in a multi-channel measurement arrangement with a number of coupling arrangements 21.

The arrangement 11 in FIG. 10 can also be set up for stimulating the biological material. At least part of the coupling surfaces and/or a circuit in the semiconductor chip 49 can be designed accordingly for this purpose. If different coupling surfaces on the respective layers 17 are used for the measurements and for the stimulation, the measurements and the stimulation can be carried out at different locations on a surface of the arrangement 11 or at different locations on the biological material. Simultaneous and spatially resolved measurement and stimulation is thus made possible. The biochip is optionally set up for simultaneous measurement and stimulation. For example, the biochip can have a number of different coupling surfaces. The biochip can have at least one coupling surface for measurement and at least one further coupling surface for simultaneous stimulation.

In summary, a biochip 13 is described here that uses the nonlinear and hysteretic curve D(F) of the electrical displacement density D in the layer with ferroelectric properties 17 as a function of the electric field E shown in FIGS To achieve charge transfer in the electrolyte 24 than would be possible with a purely capacitive stimulation. In this way, electrical stimulation of biological material can be achieved with a higher stimulation efficiency. This in turn enables the construction of microelectrode arrays and electrically active implants with a high electrode density without electrochemical charge transport between the coupling surface 15 and the electrolyte 24 for the first time a measuring device with the MOSFET 42, or with the help of a passive measuring device.

Claims

Claims Biochip (13) with at least one coupling arrangement (21) for electrically stimulating biological material (24) or for electrically measuring the biological material (24), the biochip (13) having:

- A support structure (23) which is arranged on and/or in the coupling arrangement (21); and

- a layer (17), one layer surface (19) of which is arranged on the coupling arrangement (21) and the opposite layer surface of which has a coupling surface (15) for electrical stimulation of the biological material (24) and/or for electrical measurement on the biological material (24) forms; characterized in that the layer (17) has ferroelectric properties. Biochip (13) according to claim 1, characterized in that the layer (17) has hafnium oxide with ferroelectric properties. Biochip (13) according to Claim 2, characterized in that the hafnium oxide having ferroelectric properties is doped with silicon. Biochip (13) according to Claim 2, characterized in that the hafnium oxide having ferroelectric properties is doped with zirconium. Biochip (13) according to Claim 4, characterized in that the zirconium-doped hafnium oxide with ferroelectric properties has the composition Hfo.5Zro.5O2. Biochip (13) according to claim 1, characterized in that the layer (17) has zirconium oxide with ferroelectric properties. Biochip (13) according to claim 1, characterized in that the layer (17) has aluminum scandium nitride with ferroelectric properties. Biochip (13) according to one of the preceding claims, characterized in that a thickness (d) of the layer (17) is greater than 200 nm, preferably in the range between 200 nm and 1500 nm, preferably between 200 nm and 800 nm. Biochip (13) according to one of the preceding claims, characterized in that the carrier structure (23) has a substrate (29) made of a semiconductor material. Biochip (13) according to Claim 9, characterized in that the layer (17) forms part of a ferroelectric field effect transistor (39) of the biochip (13). Biochip (13) according to claim 10, wherein the layer (17) forms an insulating layer of the ferroelectric field effect transistor (39). Biochip (13) according to one of the preceding claims, wherein the biochip (13) has memristive properties; in particular wherein the layer is formed in such a way that a ferroelectric tunnel contact is formed. Biochip according to one of the preceding claims, wherein the biochip (13) is set up to couple an artificial neural network to a biological neural network. Biochip (13) according to one of the preceding claims, characterized in that the biochip (13) has a plurality of mutually separate coupling surfaces (15) of different coupling arrangements (21) of the biochip (13). Method for producing a biochip (13) with at least one coupling arrangement (21) for electrically stimulating biological material (24) or for electrically measuring the biological material (24), the method having:

- Arranging a support structure (23) on and / or in the coupling arrangement (21); and

- Production of a layer (17) in such a way that one layer surface (19) is arranged on the coupling arrangement (21) and the opposite layer surface has a coupling surface (15) for electrical stimulation of the biological material (24) and/or for electrical measurement the biological material (24); characterized in that the layer (17) is produced as a layer with ferroelectric properties. Implant (11) with a biochip (13) with at least one coupling arrangement (21) for electrically stimulating biological material (24) or for electrically measuring the biological material (24), characterized in that the biochip is a biochip (13). any one of claims 1 to 14. Implant (11) according to claim 16, characterized in that the implant is a retina implant, a cochlear implant, an implant for deep brain stimulation and/or an implant for providing a brain-machine interface . Implant (11) according to Claim 16 or 17, characterized in that the implant has a semiconductor chip (49) separate from the biochip (13), a connection of the semiconductor chip via a conductor track (43) to an electrode layer adjoining the layer surface (19). (27) is electrically connected and the conductor track (43) and the biochip (13) are arranged on a common substrate (29). In-vitro arrangement (11) with at least one biochip (13) and an electrolyte receptacle (16) for receiving an electrolyte (24) that contains biological material, the electrolyte receptacle for arranging the electrolyte (24) on at least one coupling surface (15 ) of the biochip for electrical stimulation of the biological material and / or for electrical measurement of the biological material is set up, characterized in that the at least one biochip is a biochip (13) according to any one of claims 1 to 14. Arrangement (11) according to claim 19, characterized in that the arrangement (11) has at least one biological neuron and at least one artificial neuron.