The transformation of the analyte to the actual concentration of interest often takes

place in two or more steps in biosensors, whereby one part of the transformation can

already take place in the implant and another only after data transfer outside the body.

In the development of a Berlin-Brandenburg consortium, a microelectronic chip was

developed to exploit the transformation of glucose concentration to viscosity [7,9,27]. A

microelectromechanical system (MEMS) was developed for this purpose, in which

a mechanically deflectable beam made of titanium nitride determines the viscosity of a

mixture of ConA, glucose, and dextran [19].

Figure 21.3 shows the top view of the microelectronic chip with lateral dimensions of

1.3 × 0.4 mm, which was produced in a preparation process comprising five metal layers

and several hundred individual steps in a 0.25 µm technology [28]. The wafer was ejected

from the cleanroom in such a way that, except for the areas of the MEMS cavities to be

etched free, the top of the wafer was still covered with a photoresist. Before free etching,

however, the wafer was first thinned to 150 µm. The mechanical active structures of TiN

were exposed by wet etching of the surrounding SiO2 sacrificial layers and subsequent

critical point drying to avoid stiction (static friction) of the cantilevers. [29]. Subsequently,

the bond pads were provided with 50 µm stud bumps of gold to allow later contact of

the chip. The separation of the chips had to be done subsequently with a dry process to

avoid adhesion of the beam to the base plate, so the laser-based StealthDicing process was

used [30].

Examples of actuator systems include implants for peripheral nerve stimulation, for

which applications in pain management, cancer, and immunotherapy are envisioned,

which are sometimes referred to as electroceuticals [22]. In contrast to brain implants or

brain-computer interfaces (BCI), nerve endings on organs or tissues do not have an un­

manageable number of billions of nerve cells, but usually only a few 100 or 1,000 nerves.

Thus, there is a greater likelihood of arriving at a fundamental understanding of the effect

of electrical stimulation patterns and developing effective therapies.

Implants for vagus nerve stimulation (VNS) are already in use today and served as the

starting point for development. They have a similar design to cardio implants with a system

housing, header, and probe, except that their stimulation impulses are directed at the vagus

nerve running in the neck, which connects the brain to many internal organs. Commercially

available devices are approved for the treatment of epilepsy and depression. Other appli­

cations being studied include migraines, tinnitus, headaches, rheumatoid arthritis, hy­

pertension, and other diseases [22]. A key development step will be to reduce the size of

the electrodes, which are currently still in the mm range, but which can be manufactured

much more precisely using semiconductor technology and may address selected nerve

endings. It is expected that highly integrated CMOS technology will be used in these

systems, but the range of materials must be expanded to include materials not yet used in

semiconductor fabs, such as gold and flexible plastics like PDMS to be suitable for use in the

FIGURE 21.3

Glucose sensor chip with X-shaped mechanically bendable beams [ 9].

Implantable Microelectronics

345