Background
Recordings of electrical activity of neural cell networks can provide a wealth of information about physiology and physiological degeneration that can lead to disease, such as parkinson's disease or alzheimer's disease. MEAs have been used to monitor neural signals at regular time intervals over a long period of time by cortical electroencephalography (ECoG). This approach helps to improve researchers' insight into brain activity. Flexibility and biocompatibility are the primary conditions for MEAs to record data in the body of a test sample over an extended period of time. These properties allow for the direct placement of MEAs on the skull, in connection with the nervous system. Typically, the substrate material is selected from polyimides to provide excellent biocompatibility and high flexibility for the manufacture of MEAs. Over the past few years, flexible MEAs have been developed to stimulate and record different neurons. Researchers have now developed several high-sensitivity MEAs that use micro-electromechanical systems (MEMS), complementary metal-oxide-semiconductor (CMOS), and lab-on-a-chip micro-fabrication techniques to fabricate a large number of electrode arrays in a small scale. Furthermore, in order to be able to record a large number of individual neurons simultaneously, the electrodes are preferably operated at a high spatiotemporal resolution. In addition, precise wafer-level processes can produce MEAs with excellent uniformity, higher quality, and fault tolerance, thereby achieving high reliability. The current research focuses on the preparation of porous media by advanced microfabrication technology to reduce production cost and improve yield. Micromachining of large wafer dimensions can greatly reduce manufacturing costs due to the possibility of mass production.
Disclosure of Invention
The invention aims to solve the problems of low production efficiency and operation of flexible MEAs under high spatial and temporal resolution, and provides a cortical electroencephalogram-based multi-channel MEAs wafer-level manufacturing and assembling method.
The preparation method of the multichannel MEAs based on the cortical electroencephalogram is realized according to the following steps:
firstly, immersing a silicon substrate into an HF solution to remove an oxide layer, and cleaning the silicon substrate by using plasma to obtain a cleaned silicon substrate;
secondly, depositing an aluminum sacrificial layer on the surface of the cleaned silicon substrate by an electron beam process;
thirdly, spin-coating a non-photosensitive polyimide (MEAs) base material on the aluminum sacrificial layer for multiple times, and sequentially carrying out soft baking and curing to obtain a substrate with a polyimide bottom layer with the thickness of 8-10 microns;
fourthly, patterning and shaping the polyimide layer by adopting a photoetching process, wherein the cavity pressure is 440-460 mTorr, and the power is 620-680W, CF4The flow rate is 6-10 sccm and O2Carrying out inductively coupled plasma etching treatment on the polyimide layer under the condition that the flow is 70-80 sccm to obtain a patterned substrate;
forming a multichannel micro (recording) electrode network and Interconnection boards (Interconnection Pads) by a photoetching process, wherein the micro (recording) electrode is made of a chromium/platinum metal layer, and removing redundant metal layers by a stripping process to obtain a wafer with the microelectrode network;
spin-coating a non-photosensitive polyimide (MEAs) substrate on a polyimide bottom layer, sequentially carrying out soft baking and curing to obtain top layer polyacetamide with the thickness of 8-10 micrometers, masking the wafer by using a photoresist to limit an interconnection line and a bonding pad, then forming a chromium/platinum metal layer by using an electron beam evaporator, and stripping the masks of the interconnection line and the bonding pad;
and seventhly, removing the aluminum sacrificial layer by adopting an aluminum etchant wet etching process to obtain the multichannel MEAs based on the cortical electroencephalogram.
The invention relates to a cortical electroencephalogram-based multichannel MEAs, wherein a microelectrode array is arranged between two layers of flexible polyimide substrates, and the arrangement mode of the microelectrode array is as follows: multiple rows of microelectrodes are arranged in parallel at intervals, the microelectrode units in each row of microelectrodes are arranged at intervals, the lead of each microelectrode unit is led out from the middle part of each row of microelectrodes, and the lead is connected with the interconnection board.
In the invention, a 6-inch high-resistance silicon wafer is used as a substrate, and 25 MEAs can be produced on a single wafer in large quantity by combining flexible polyimide. The recording locations are compactly arranged with high spatial and temporal resolution. By optimizing the MEAs structure, the interference between adjacent electrodes can be effectively reduced. In addition, interconnect pads have been added in particular in the proposed work to connect the slot connectors and the devices being fabricated. The maximum utilization achievable after consideration of wafer edge defects and manufacturing tolerance factors is achieved. Surface mount devices are manufactured with slot connectors without any additional bond wires. This configuration prevents electrical interconnection failure of the device, which may be caused by cranial tissue during in vivo testing. Flexible printed circuit boards (FPCB, polyimide based structures) are used to support MEAs and provide protection from physical damage. The top and bottom PCBs are formed with connector sockets on the front and interconnect pads on the back. The FPCB is connected to the top and bottom PCBs through Vcut, which can be easily removed during in vivo testing. In general, the device can record cranial signals for long periods of time, preferably several rounds of recording to provide accurate and reliable measurements. Thus, the top and bottom PCBs and the slot connector are designed to be detachable and may be detached during in vivo recording. This innovative circuit helps to alleviate the pain of the patient due to the extra weight of the top and bottom PCBs. In addition, scalp operation is not needed to be performed again in the next round of measurement, and the method is favorable for improving the satisfaction degree of a patient while saving cost and reducing complexity.
10 MEAs samples were evaluated by electrochemical impedance spectroscopy. Impedance spectroscopy proves that the assembled MEAs have good stability and can be used for simultaneously recording a neuron network with high selectivity and high sensitivity of a plurality of neurons. Finally, the cranium of adult male mice was tested in vivo. High reliability and excellent yield can significantly reduce the price of MEAs and provide opportunities for the future biomedical market to pursue commercial success.
Drawings
FIG. 1 shows the silicon wafer passing through O in the example2/H2A surface topography map after plasma treatment;
FIG. 2 shows the polyimide wafer diameter O in the example2A surface profile after plasma treatment;
FIG. 3 is a schematic diagram of the structure of a cortical electroencephalogram based multichannel MEAs, 1-microelectrode, 2-interconnection plate, 3-interconnection;
FIG. 4 is a block diagram of the dimensions of a single cortical electroencephalogram-based multichannel MEAs;
FIG. 5 is a pictorial view of a single cortical electroencephalogram based multichannel MEAs;
FIG. 6 is a pictorial representation of 25 MEAs on a fully processed 6 inch silicon wafer;
FIG. 7 is a block diagram of the connection of cortical electroencephalogram based multichannel MEAs to a connector; the PCB comprises 4-FPCB, 5-top PCB, 6-bottom PCB, 7-V-cut and 8-connector;
FIG. 8 is a top view of cortical electroencephalogram based multichannel MEAs mounted on a PCB surface;
FIG. 9 is a test chart of measured impedance characteristics of 10 random channels in a cortical electroencephalogram-based multichannel MEAs;
FIG. 10 is a graph of ECoG channel measurements taken with 60 channels per MEAs in an example;
FIG. 11 is an image of flexible MEAs mounted on the skull of an adult male mouse in an example;
FIG. 12 is an electroencephalogram sample measurement of an intraperitoneal injection of GBL causing paroxysmal epilepsy;
FIG. 13 is a microelectrode distribution diagram of the cortical electroencephalogram-based multichannel MEAs of the example.
Detailed Description
The first embodiment is as follows: the preparation method of the multichannel MEAs based on the cortical electroencephalogram is implemented according to the following steps:
firstly, immersing a silicon substrate into an HF solution to remove an oxide layer, and cleaning the silicon substrate by using plasma to obtain a cleaned silicon substrate;
secondly, depositing an aluminum sacrificial layer on the surface of the cleaned silicon substrate by an electron beam process;
thirdly, spin-coating a non-photosensitive polyimide (MEAs) base material on the aluminum sacrificial layer for multiple times, and sequentially carrying out soft baking and curing to obtain a substrate with a polyimide bottom layer with the thickness of 8-10 microns;
fourthly, patterning and shaping the polyimide layer by adopting a photoetching process, wherein the cavity pressure is 440-460 mTorr, and the power is 620-680W, CF4The flow rate is 6-10 sccm and O2Carrying out inductively coupled plasma etching treatment on the polyimide layer under the condition that the flow is 70-80 sccm to obtain a patterned substrate;
forming a multichannel micro (recording) electrode network and Interconnection boards (Interconnection Pads) by a photoetching process, wherein the micro (recording) electrode is made of a chromium/platinum metal layer, and removing redundant metal layers by a stripping process to obtain a wafer with the microelectrode network;
spin-coating a non-photosensitive polyimide (MEAs) substrate on a polyimide bottom layer, sequentially carrying out soft baking and curing to obtain top layer polyacetamide with the thickness of 8-10 micrometers, masking the wafer by using a photoresist to limit an interconnection line and a bonding pad, then forming a chromium/platinum metal layer by using an electron beam evaporator, and stripping the masks of the interconnection line and the bonding pad;
and seventhly, removing the aluminum sacrificial layer by adopting an aluminum etchant wet etching process to obtain the multichannel MEAs based on the cortical electroencephalogram.
In the fourth step of the present embodiment, the RMS of the polyimide wafer after the etching treatment with the inductively coupled plasma is 14 to 14.2 nm.
The second embodiment is as follows: the present embodiment is different from the first embodiment in that the mass concentration of the HF solution in the first step is 10%.
The third concrete implementation mode: the difference between the present embodiment and the first embodiment is that the plasma cleaning process in the first step is as follows:
controlling O in a plasma cleaner2/H2The flow ratio of (2) is 9000: at 450sccm, clean for 30 seconds at a RF power of 650W, a substrate (chuck) temperature of 80 deg.C, and a chamber pressure of 2 Torr.
The fourth concrete implementation mode: the difference between the first embodiment and the second embodiment is that the root mean square value (RMS) of the roughness of the surface of the silicon substrate in the first step is 8.5-8.8 nm.
Controlling the surface roughness of the silicon substrate in this embodiment helps to enhance the adhesion between the silicon substrate and the sacrificial layer of aluminum.
The fifth concrete implementation mode: the difference between this embodiment and one of the first to fourth embodiments is that the sacrificial layer of aluminum in step two has a thickness of 3 μm.
The sixth specific implementation mode: this embodiment is different from one of the first to fifth embodiments in that the soft baking in the third and sixth steps is performed at 100 ℃ for 3 minutes, and the curing in the third and sixth steps is performed at 300 ℃ for 3 minutes.
The seventh embodiment: this embodiment is different from one of the first to sixth embodiments in that the thickness of the polyimide layer in the third step and the thickness of the polyimide layer in the sixth step are both 9 μm.
The specific implementation mode is eight: the difference between this embodiment and the first to seventh embodiments is that the arrangement of the microelectrode network in the fifth step is as follows: multiple rows of microelectrodes are arranged in parallel at intervals, the microelectrode units in each row of microelectrodes are arranged at intervals, and the lead of each microelectrode unit is led out from the middle part of each row of microelectrodes.
The specific implementation method nine: the difference between this embodiment and the eighth embodiment is that the microelectrode network comprises 50 to 70 microelectrode units, and 10 to 12 rows of microelectrodes are arranged in parallel at intervals.
The microelectrode network of the embodiment has a Christmas tree structure, and the FPCB of the MEAs extends out of the mouse body through the Vcut and is connected with the PCBs at the top and the bottom.
The detailed implementation mode is ten: this embodiment differs from one of the first to ninth embodiments in that the deposition rates of chromium and platinum in the chromium/platinum metal layer in step five are both 3 angstroms/second.
Example (b): the preparation method of the multichannel MEAs based on the cortical electroencephalogram is implemented according to the following steps:
firstly, a silicon substrate is immersed into 10% HF solution to remove an oxide layer, and O is controlled in a Plasma Cleaner (RF Plasma Cleaner)2/H2The flow ratio of (2) is 9000: plasma cleaning at a radio frequency of 650W, a chuck temperature of 80 ℃ and a chamber pressure of 2Torr for 30 seconds at a temperature of 450sccm to obtain a cleaned silicon substrate, wherein the root mean square value (RMS) of the base roughness is 8.64nm, and the accurate surface roughness is helpful for enhancing the adhesion between the silicon substrate and the aluminum sacrificial layer;
secondly, depositing an aluminum sacrificial layer with the thickness of 3 microns on the surface of the cleaned silicon substrate through an electron beam process;
thirdly, spin-coating a non-photosensitive polyimide MEAs substrate (HD Microsystems, PIX1400) on the sacrificial layer of aluminum three times, depositing the polyimide substrate by a spin coater three times at 500/3000/500 rpm and 10/40/5s for 3 microns each time, soft baking at 100 ℃ for 3 minutes, and curing at 300 ℃ for 3 minutes in an oven to obtain a substrate with a 9 micron thick polyimide layer, which can relieve the tensile stress generated in the upcoming metallization process to overcome the potential rolling problem and planarize the film, however, if the thickness of the polyimide film exceeds 9 microns, the film will have the following blocking problem when attached to the skull of a mouse;
patterning and shaping the polyimide layer by adopting a photoetching process, treating at 90 ℃ for 30min, then treating at 125 ℃ for 60min, and curing to obtain stable height and shape with excellent aspect ratio, and then keeping the cavity pressure at 450mTorr and the power at 650W, CF4The flow rate is 8sccm and O2Carrying out inductively coupled plasma etching treatment on the polyimide layer under the condition that the flow rate is 72sccm, promoting the adhesion between the first layer of polyimide and the Cr/Pt metal layer as well as between the first layer of polyimide and the second layer of polyimide, wherein the root mean square value is 14.01nm, and obtaining a patterned substrate;
forming 60-channel microelectrode network on the patterned substrate by photoetching, wherein the chromium/platinum (15/150 nm) metal layer is evaporated at a proper deposition rate of 3A/s and 3A/s, then performing a stripping process, wherein the device is cleaned by using 3 MPa pressure, treated by acetone for 60 s, treated by isopropanol for 30 s, and finally treated by deionized water for 60 s;
spin-coating the non-photosensitive polyimide MEAs substrate on the polyimide bottom layer again, sequentially carrying out soft baking and curing to obtain top layer polyacetamide with the thickness of 9 microns, masking the wafer by using a photoresist to limit an interconnection line and a bonding pad, then forming a chromium/platinum metal layer by using an electron beam evaporator, wherein the deposition rates are 3 angstroms/second and 3 angstroms/second respectively, and stripping the masks of the interconnection line and the bonding pad;
and seventhly, removing the aluminum sacrificial layer by adopting an aluminum etchant wet etching process to obtain the multichannel flexible MEAs based on the cortical electroencephalogram.
The microelectrode network described in the fifth step of this embodiment is composed of 60 channels of microelectrodes, the microelectrode array is in a shape of "christmas tree" (as shown in fig. 4), 12 rows of microelectrodes are arranged in parallel at intervals, the microelectrode units in each row of microelectrodes are arranged at intervals, the lead of each microelectrode unit is led out from the middle part of each row of microelectrodes, and the lead is connected with the interconnection board. The FPCB of the "christmas tree" architecture MEAs extend out of the mouse body through Vcut and connect to the top and bottom PCBs (as shown in fig. 7 and 8).
This example prepares flexible polyimide-based MEAs with 60 channels. The flexible FPCB and the slot connector are applied to provide surface mount assembly without requiring an additional wire bonding process and provide an easier signal recording method for manufactured MEAs. 10 MEAs samples were evaluated by electrochemical impedance spectroscopy. Impedance proves that the assembled MEAs have good robustness and can be used for simultaneously recording a neuron network with high selectivity and high sensitivity of a plurality of neurons. High reliability and excellent yield can significantly reduce the price and provide opportunities for MEAs to pursue commercial success in the biomedical market in the future. Finally, the skull of the adult male mouse is subjected to in vivo test, and the skull structure is proved to be very suitable for signal recording by the method.