Microsystems for biomedical applications are getting a lot of attention last decade. Such microsystems consist of one or more chips, sensors, a battery, electrodes, etc. They are developed to be used as wearable are implantable device, increasing the health of a patient or to support a patient with a disease or failing organ. Imec is developing since many years electronic silicon probes with miniature electrodes for recording the electric activity of neural cells, in order to enable the medical world to improve their neural research. Some devices contain also electrodes for stimulation of neural cells by enabling local charge transfer into the cells. As such, imec was/is playing a major role in multi-partner international projects such as Neuropixels, Brainstar, Neuroseeker, etc. (See Fig. 1).
Fig. 1: a Silicon based electronic device for brain investigation, developed at imec in the frame of the international multi-partner project Neuroseeker. The Silicon chip has a 8mm long and 50um thin Si needle containing a huge amount of miniaturized TiN electrodes, in order to enable brain research by recording the neural activity in the brain. The device is developed for brain research purpose, the current realization is not suited for long term human usage.
In order to record electrical signals from neural cells, electrodes need to have a good contact with the local tissue in order to realize a low electrode/tissue impedance. When electrical stimulation has to be performed, the requirements for the electrodes are more stringent, since it is not easy to transfer the electric current from the electrode into the local tissue. This property is related to the so called the 'Charge Storage Capacity' (CSC) and 'Charge Injection Capacity' (CIC) of an electrode and both are important properties for stimulation electrodes. The easiest way to increase the CSC and CIC is to enlarge the electrode surface, but obviously this is in conflict with the need of miniaturization of the implanted device. Hence research is needed towards alternative electrode materials, or towards optimization of electrode fabrication, in order to improve the CSC/CIC and other important properties. While gold and platinum are interesting biocompatible electrode materials for recording electrodes, their CSC/CIC is not very high. Imec started to investigate alternative stimulation electrode materials compatible with CMOS processing, and obtained already interesting results with optimized TiN electrodes (See Fig. 1). These material investigations however have been focusing on short term usage of electrodes for research purposes only, not for long term implantation in humans. Therefore imec wants to expand its research towards alternative electrode materials for stimulation, with special attention for long term electrode functionality and biostability. In order to predict long term behavior, various types of accelerated testing protocols have to be performed/ developed (both thermal and chemical acceleration) during this PhD project. Various biocompatible conductive materials can be tested, such as platinum (as a reference), Iridium oxide, TiN, Ruthenium, etc.
Since long term implantation in humans is the final goal of electrical stimulation devices, the total implanted device itself as well as the electrodes have to fulfil dedicated requirements next to its functionality, such as biocompatibility, biostability and device hermeticity. Indeed, in case of in-vivo use of electronics, a lot of extra measures have to be taken into account: the body has to carry the implant without harmful consequences, and moisture from the body should not penetrate into the implant, leading to corrosion resulting in failure of the microsystem. Hence a biocompatible and hermetic package is needed for such implantable systems. Traditionally, an implantable system is placed into a rigid Titanium box (See Fig. 2). Such box is indeed biocompatible and hermetic, offering the 'safety' of a well-known package for implants, but at the high price of large implantable devices with more patient discomfort and a higher risk on complications after implantation, such as a pronounced foreign body reaction or infections.
Fig 2. Electronic chip for neural implantation, fully packaged by CMST/imec's ultrathin flexible hermetic encapsulation based on the FITEP technology platform. The total packaged chip is only 75nm thick, allowing for a minimum invasive implantation into the peripheral nerves, with minimal tissue damage and hence reduced scar tissue formation as important advantage. The electrodes for neural recording and stimulation are located on top of the chip. (This first FITEP demonstrator was realised during the IMPRESS project, which was sponsored by DARPA BTO under the auspices of Dr. D. Weber through SPAWAR, Pacific Grant/Contract No. N66001-15-C-4018 to the University of Florida. )
Within a collaboration between CMST (an Imec-associated lab at the UGent university) and Imec in Leuven, a considerable amount of research is performed within our so called FITEP technology platform (FITEP: Flexible Implantable Thin Electronic Package). The goal is to develop an advanced implantable packaging technology, in order to realize a very small, flexible, biomimetic package for electronic implants. This implantable package should result in excellent device functionality, patient safety and comfort. Moreover, due to the strong miniaturization and flexible character of the novel package, the technology would open important new possibilities to the medical world. First interesting results are obtained already by packaging a CMOS chip for neural implantation, as can be seen in Fig. 2. The chip contains many electrodes for recording as well as for the stimulation of the neural cells close to the device. The device encapsulation is the result of dedicated research at CMST and consists of a multilayer of biocompatible polymers and ultrathin ceramic diffusion barriers deposited using the ALD technique (ALD: atomic layer deposition) in order to fabricate a very thin and flexible but also highly hermetic device packaging. This polymer/ALD stack proved to be a very hermetic enclosure: copper patterns protected with the polymer/ALD stack are still in perfect condition after more than 2 years of immersion in saline at 60°C, while Cu patterns protected by the polymer stack without ALD barriers showed first signs of damage already after 6 weeks exposure to saline. This thermal acceleration test is still ongoing, hopefully for many more years. Hence we proved that out optimized barrier is functional as protective barrier for at least 10 years in saline at 37°C. Very important however is to remark that the barrier is indeed excellent, but without holes etched through it in order to allow direct contact between electrodes on the device and the saline. When a real device is fabricated, the barrier encapsulation will have etched windows on the location of the electrodes, and integration technologies will have to be developed to ensure no leakage pads are created by opening windows in the encapsulating diffusion barrier. For our first FITEP demonstrator the electrode windows are realized by a multitude of polymer and ALD layers overlapping with platinum electrodes, as can be seen in Fig. 3. Only short term functionality is proven yet, for long term stability several improvements of the structure are essential. Although impressive results are obtained already with our first demonstrator, we would like to expand our excellence in this field by this PhD research project, which will have 2 strongly related goals. Hence the PhD student will focus on both, optimization of electrode materials for superior long term stimulation functionality, as well as on integration of these optimized electrodes in the biocompatible hermetic encapsulation without damaging the encapsulation quality.
Fig. 3: First FITEP demonstrator: cross-section of the Si-chip encapsulated with the polymer/ALD multi-stacks. During and after the encapsulation with these multi-stacks, windows are etched into the encapsulation layers to open the platinum electrodes. Finally an Iridium oxide layer is deposited on the electrodes using lift-off, in order to improve the stimulation properties. Short term functionality is proven, but for long term stability several improvements of the structure are essential.
This research will be carried out on two locations: at CMST in Ghent supervised by the PhD promotor, and at imec in Leuven under supervision of the co-promotor. As such the PhD student can make optimal use of the available knowledge and infrastructure at both locations. The work consists of processing work in the cleanroom, as well as performing measurements in the related biolab and reliability lab. Furthermore, the development of a dedicated test tool and protocol is envisaged in order to enable accelerated chemical testing. Obviously, general research planning and follow-up during the PhD project will be in consultation with both involved research teams at CMST and at imec.
For more information: see following publications
- Op de Beeck Maaike et al., "Ultra-thin biocompatible implantable chip for bidirectional communication with peripheral nerves", proceedings IEEE BIOCAS Conf., Oct. 2017, Torino, Italy; DOI: 10.1109/BIOCAS.2017.8325206
- James J. Jun, et al., "Fully integrated silicon probes for high-density recording of neural activity", Nature volume 551, pages 232–236 (09 November 2017) https://www.nature.com/articles/nature24636
Required background: Engineering technology, Engineering science, Biomedical engineering, Chemical engineering, or equivalent
Type of work: 60% experimental, 30% interpretation, 10% theory/ literature study
Supervisor: Maaike Op de Beeck, Liesbet Lagae
Daily advisor: Maarten Cauwe, Dries Braeken
The reference code for this position is 1812-54. Mention this reference code on your application form.