Elsevier

Solid-State Electronics

Volume 52, Issue 9, September 2008, Pages 1364-1373
Solid-State Electronics

Joining microelectronics and microionics: Nerve cells and brain tissue on semiconductor chips

https://doi.org/10.1016/j.sse.2008.04.024Get rights and content

Abstract

The direct electrical interfacing of semiconductor chips with individual nerve cells and with brain tissue is considered. At first, the structure of the cell–chip contact is described and then the electrical coupling is characterized between ion channels, the electrical elements of nerve cells, and transistors and capacitors of silicon chips. On that basis, the signal transmission between microelectronics and microionics is implemented in both directions. Simple hybrid systems are assembled with neuron pairs and with small neuronal networks. Finally, the interfacing with capacitors and transistors is extended to brain tissue on silicon. The application of CMOS chips with capacitively coupled recording sites allows an imaging of neuronal activity with high spatiotemporal resolution. Goal of the work is an integration of neuronal network dynamics and digital electronics on a microscopic level for applications in brain research, medical prosthetics and information technology.

Introduction

Computers and brains work electrically. However, their charge carriers are different, electrons in solid silicon and ions in liquid water. It is an intellectual and technological challenge to join these different systems right on the level of their electronic and ionic signals. In principle, Luigi Galvani has established an electrical coupling between inorganic solids and excitable living tissue in the 18th century. Today, after fifty years of dramatic developments in semiconductor technology as well as in cellular neurobiology, we may envisage a most complex integration of microionics and microelectronics with myriads of nerve cells and semiconductor devices and breathtaking applications in medicine and information technology.

Usually, in neurophysiology the coupling of electronic and ionic signals is achieved with perfectly unpolarized solid/water contacts such as Ag/AgCl electrodes where ionic and electronic currents are transformed into each other. That approach is not suitable for an iono-electronic interfacing on a microscopic scale with numerous contacts. Semiconductor chips must be protected from corrosion, nerve cells from electrochemical reaction products. Interfacing must be implemented without Faradaic current. The communication between microionics and microelectronics must be achieved by displacement currents across an insulating interface [1], [2].

The present paper describes step by step the mechanism of neuroelectronic interfacing on the scale of nanometers, micrometers and millimeters. It starts with the contact between an individual nerve cell and a silicon chip. Then the interfacing of the fundamental devices of the brain and the fundamental devices of computers is studied. On that basis, the interfacing of nerve cells with capacitors and transistors and the assembly of simple neuroelectronic hybrids with two nerve cells and with simple neuronal networks is considered. Finally, the challenges with electronic interfacing of brain tissue are addressed.

Section snippets

Cell–chip contact

A simple hybrid with a nerve cell from rat brain and a sensor transistor in silicon is depicted in Fig. 1. The cell is surrounded by a membrane with an electrically insulating core of lipid. That lipid bilayer (thickness about 5 nm) separates the bath with a concentration of 150 mM sodium ions from the cytoplasm with about 150 mM potassium ions. The silicon is coated with thermally grown silicon dioxide (thickness 10 nm). The basic problem with respect to the electrical interaction of cell and chip

Ion–electron coupling

The electrical elements of nerve cells are voltage-gated ion channels. These molecules are embedded in the lipid bilayer of the membrane. They can be in an open and in a closed state. When they are open, they selectively transmit ionic current through the membrane, Na+ inward current or K+ outward current. The opening and closing of the channels is connected with a displacement of electrical charge across the membrane. As a consequence, opening and closing can be controlled by the voltage

Semiconductor chips with nerve cells

A first step towards an integration of neuronal dynamics and microelectronics is the interfacing of individual nerve cells with silicon microstructures. We consider the excitation of nerve cells by EOS capacitors and the recording of neuronal activity by EOSFETs.

A crucial issue is the contribution of two domains of a nerve cell, the membrane that is attached to the chip and the upper free membrane in contact to the bath as illustrated in Fig. 6: (i) When a voltage ramp is applied to a

Elementary neuroelectronic hybrids

In a first step towards the assembly of neuroelectronic hybrids, we couple pairs of nerve cells to a chip with two different pathways as illustrated in Fig. 12.

Neuronal networks on chip

The function of complex neuronal networks relies (i) on an ordered mapping between sets of neurons by synaptic connections and (ii) on an enhancement of the synaptic strength by correlated presynaptic and postsynaptic activity (Hebbian learning). An experimental study of that function requires (i) neuronal nets with a defined topology of the synaptic connections, and (ii) a noninvasive supervision of all neurons to induce learning and to observe the performance of the network. To achieve these

Semiconductor chips with brain tissue

Culturing of defined neuronal networks can be avoided when neuronal networks are used that are provided by brains. Brain tissue with a planar structure of the networks is required to attain an efficient interfacing with a planar chip. Organotypic brain slices are particularly promising because they are a few cell layers thick and conserve major neuronal connections. Compared to the culture of dissociated cells, however, there are new problems if we want to interface individual nerve cells in a

Outlook

The basic problems on the electrical interfacing of individual nerve cells and semiconductor chips are solved. Important progress was achieved by the development of stimulation EOS capacitors with high dielectric constant, of sensor EOS transistors with low-noise and of CMOS chips with a high density of sensor sites.

With respect to hybrid systems of neuronal networks and digital microelectronics, we are still in an elementary stage. Several directions may be considered: (i) small defined

References (26)

  • B. Kuhn et al.

    Anellated hemicyanine dyes in a neuron membrane: molecular Stark effect and voltage recording

    J Phys Chem

    (2003)
  • F. Wallrapp et al.

    TiO2 and HfO2 in electrolyte–oxide–silicon configuration for applications in bioelectronics

    J Appl Phys

    (2006)
  • M. Ulbrich et al.

    Opening of K+ channels by capacitive stimulation from silicon chip

    Appl Phys A

    (2005)
  • Cited by (34)

    • Towards Oxide Electronics: a Roadmap

      2019, Applied Surface Science
    • Wireless system for biological signal recording with Gallium Arsenide high electron mobility transistors as sensing elements

      2013, Microelectronic Engineering
      Citation Excerpt :

      We here propose the use of Gallium Arsenide (GaAs) – based high electron mobility transistors (HEMTs) connected to a wireless link to realize a novel biotelemetry sensor with several advantages as compared to diodes based approaches. Field effect transistors (FETs) analogous to HEMTs have been largely employed as sensors for the detection of biological activity, ranging from pH variations to macromolecules sensing and to electrical activity from cultured neuronal networks [8–10]. Moreover, FETs fabrication technology is nowadays extremely advanced, allowing implementation on very small areas with unvaried performances and reduced noise figures as compared to miniaturized diode mixers.

    • Gallium arsenide passivation method for the employment of High Electron Mobility Transistors in liquid environment

      2012, Microelectronic Engineering
      Citation Excerpt :

      Most of the work with FET biochemical sensors exploits silicon technology. In particular, interfacing of silicon chips with nerve cell and brain tissue has been demonstrated [2], and advances in technology are opening the way to in vivo applications [3]. However, more performing devices, with higher sensitivity and faster response, could be realized by exploiting different materials.

    • Atomic layer deposited TiO<inf>2</inf> for implantable brain-chip interfacing devices

      2012, Thin Solid Films
      Citation Excerpt :

      Fig. 4 illustrates the C–V curves at 1 KHz for 11.5 nm thick TiO2/1 nm SiOx-like IL/pSi and 12.5 nm TiO2/1 nm Al2O3/pSi stacks respectively. The k-value of both stacks including TiO2, as extracted from the accumulation capacitance value, results from 5 to 6 times larger compared to that of an EOS junction including SiO2 as dielectric layer, making the system suitable for efficient capacitive neuron–silicon coupling [5]. In addition, the presence of the alumina buffer layer further increases the specific capacitance of the system, to 1.64 μF/cm2 respect to 1.49 μF/cm2 for TiO2 on Si with no buffer layer.

    View all citing articles on Scopus
    View full text