Search

Objective. Flexible neural probes are hypothesized to reduce the chronic foreign body response (FBR) mainly by reducing the strain-stress caused by an interplay between the tethered probe and the brain's micromotion. However, a large discrepancy of Young's modulus still exists (3-6 orders of magnitude) between the flexible probes and the brain tissue. This raises the question of whether we need to bridge this gap; would increasing the probe flexibility proportionally reduce the FBR? Approach. Using novel off-stoichiometry thiol-enes-epoxy (OSTE+) polymer probes developed in our previous work, we quantitatively evaluated the FBR to four types of probes with different softness: silicon (∼150 GPa), polyimide (1.5 GPa), OSTE+Hard (300 MPa), and OSTE+Soft (6 MPa). Main results. We observed a significant reduction in the fluorescence intensity of biomarkers for activated microglia/macrophages and blood-brain barrier (BBB) leakiness around the three soft polymer probes compared to the silicon probe, both at 4 weeks and 8 weeks post-implantation. However, we did not observe any consistent differences in the biomarkers among the polymer probes. Significance. The results suggest that the mechanical compliance of neural probes can mediate the degree of FBR, but its impact diminishes after a hypothetical threshold level. This infers that resolving the mechanical mismatch alone has a limited effect on improving the lifetime of neural implants. (Less)

This paper describes the fabrication and initial characterization of an ultra-thin silicon PIN detector using a new technique in silicon nanotechnology. In collaboration with the Nuclear Physics Division and the Lund Nano Lab at Lund University, we have developed and manufactured ultra thin ΔE-detectors for spectroscopic applications. The fabrication process has been carried out using a double-polished silicon substrate n-type wafer and locally thinning by means of a 10:1 solution of 25% tetramethyl ammonium hydroxide (TMAH) with Isopropyl alcohol. More than 100 detectors of different thicknesses, down to 5 μm with active areas ranging from 0.71 to 0.172 mm2, have been fabricated. The main design considerations of our thin detectors were a very low leakage current below 12 nA and a low full depletion voltage at a reverse bias less than 1.5 V. Finally, most of our thin detectors offer an energy resolution (FWHM) as low as 31 keV for 5.487 MeV alpha particles from a 241Am source.

We compared porous silicon (pSi) with smooth Si as chip-implant surfaces in a nerve regeneration setting. Silicon chips can be used for recording neural activity and are potential nerve interface devices. A silicon chip with one smooth and one porous side inserted into a tube was used to bridge a 5 mm defect in rat sciatic nerve. Six or 12 weeks later, new nerve structures surrounded by a perineurium-like capsule had formed on each side of the chip. The number of regenerated nerve fibers did not differ on either side of the chip as shown by immunostaining for neurofilaments. However, the capsule that had formed in contact with the chip was significantly thinner on the porous side than on the smooth side. Cellular protrusions had formed on the pSi side and the regenerated nerve tissue was found to attach firmly to this surface, while the tissue was hardly attached to the smooth silicon surface. We conclude that a pSi surface, due to its large surface area, diminished inflammatory response and firm adhesion to the tissue, should be a good material for the development of new implantable electronic nerve devices.

Axonal growth and guidance were studied on different micro- and nanostructured surfaces. Nanoimprinted grooves in a polymer, epitaxial III/V nanowires, porous silicon patterns, and chemically altered surfaces were all shown to induce axonal guidance. Neurons were also found to be able to attach and grow on gallium phosphide nanowires without compromising cell survival. The results are important for the construction of a new generation of neuroelectrical interfaces, including high spatial resolution electrodes. The advantages of the different nanostructured surfaces are discussed.

A capillary filling microfluidic proteomic sample processing system has been realized in an automated set-up. Chip-integrated solid-phase extraction is performed in a 96-array format followed by sequential capillary action elution into a piezoelectric microdispenser and subsequent transfer to MALDI targets. Samples are eluted and deposited in volumes of 200–300 nl. The robotic system offers calibration to user-defined microextraction arrays and MALDI-target formats at micrometre resolution. Built-in force feedback control ensures a precise and robust microchip docking/handling in three dimensions. An efficient automated washing protocol eliminates analyte carry-over. System throughput ranges typically from 50–100 samples h-1.

The paper studies the conditions optimum for the proliferation of neural cells at the surface of porous silicon. The results of this study show that the best results are achieved with meso-porous silicon. Geometrical properties of the wafers are also important through charge distribution at mask edges, change of hydrophobicity, degree of oxidation, the release of adhesion proteins from regenerating axons, etc

A new beam blanking system was installed at the Lund Nuclear Microprobe and employed in proton beam lithography (PBL) for polymer microstructures fabrication. The blanker consists of two parallel plates connected to a high voltage generator. Measurement of the beam blanking time on a sample was performed by means of the standard PIXE system. The beam is blanked and returns to a sample within 200 ns. The blanking system is designed for the new sub-micrometer beamline under installation in the accelerator laboratory. A number of pilot MeV ion beam lithography experiments were performed to illustrate the possibility to use the blanking system in combination with the existing data acquisition and scanning system. A 2.5 MeV proton beam was used to irradiate 50 mum SU-8 negative resist. The blanker was shown to be a necessary part of the lithography system. It enables blanking between each pixel and hence fabrication of various patterns down to a single pixel. The blanker has significantly simplified beam control and enhanced process time and spatial resolution. Three-dimensional microstructures with 20:1 aspect ratio were fabricated.

In this study the field effects on electrochemically controlled silicon etching in KOH during three- and four-electrode pn etch-stop have been investigated. When a p-doped bulk area is surrounded by n-doped silicon and the n-doped silicon is electrochemically passivated, small lines of bulk silicon in between the n-doped silicon would not etch. The size of the bulk silicon that did not etch could be moderated by the applied potential on the n-doped silicon during pn etching. For the three-electrode system the size of the electric field passivated bulk silicon could be controlled for line widths ranging between 60 and 100 μm. When a four-electrode system was used and the p-doped bulk silicon was forced to a more negative potential, it was possible to etch lines of bulk silicon down to 40 μm widths. This field-restricted pn etch-stop can be used for fabricating thin membrane structures with densely spaced n-doped regions that can be individually addressed by electronics. One application is the fabrication of neural sieve electrodes which comprise a pn etch-stopped membrane containing several phosphorous-doped individually addressable recording electrodes.

A neural interface could be used to control a limb prosthesis. Such an interface can be created by facilitating axonal regeneration through a sieve electrode and then register nerve signals intended to control the prosthesis. A key question is how to design the electrodes to ensure the best possible regeneration. Our previous studies have indicated that regeneration can be achieved using electrodes with square-shaped, 100 x 100 microm, via holes (holes that axons will regenerate through). Other reports have indicated a suitable range of these holes between 40 and 65 microm. In the present study we used silicon sieve electrodes with via holes of either 30 or 90 microm. The transparency, i.e. the percentage of the total via hole area, of these electrodes was either 20 or 30%. The electrodes were inserted into a silicone chamber which was used to bridge a gap in a rat sciatic nerve. After 12 weeks of nerve regeneration electrodes with a hole size of 30 microm and a 30% transparency had the most favourable result as judged by the regained gastrocnemius muscle force and the formation of reactive tissue inside the chamber. The sieve electrode transparency is crucial for ensuring regeneration.

We have developed a multichannel electrode array—termed \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{upgreek} \usepackage{mathrsfs} \setlength{\oddsidemargin}{-69pt} \begin{document} }{}\(\mu \) \end{document}-foil—that comprises ultrathin and flexible electrodes protruding from a thin foil at fixed distances. In addition to allowing some of the active sites to reach less compromised tissue, the barb-like protrusions that also serves the purpose of anchoring the electrode array into the tissue. This paper is an early evaluation of technical aspects and performance of this electrode array in acute in vitro/in vivo experiments. The interface impedance was reduced by up to two decades by electroplating the active sites with platinum black. The platinum black also allowed for a reduced phase lag for higher frequency components. The distance between the protrusions of the electrode array was tailored to match the architecture of the rat cerebral cortex. In vivo acute measurements confirmed a high signal-to-noise ratio for the neural recordings, and no significant crosstalk between recording channels., We have developed a multichannel electrode array, termed m-foil, that comprises ultrathin and flexible electrodes protruding from a thin foil at fixed distances. This study is an early evaluation of technical aspects and performance of this electrode array in acute in vitro/in vivo experiments. The in vivo acute measurements showed a good signal-to-noise ratio for the neural recordings, and no significant cross-talk between recording channels.

We present an electrode, based on structurally controlled nanowires, as a first step towards developing a useful nanostructured device for neurophysiological measurements in vivo. The sensing part of the electrode is made of a metal film deposited on top of an array of epitaxially grown gallium phosphide nanowires. We achieved the first functional testing of the nanowire-based electrode by performing acute in vivo recordings in the rat cerebral cortex and withstanding multiple brain implantations. Due to the controllable geometry of the nanowires, this type of electrode can be used as a model system for further analysis of the functional properties of nanostructured neuronal interfaces in vivo.

A polymer foil based array with 9 gold electrodes for chronic electrophysiological recordings in the CNS has been developed. A polymer, SU-8, is pattered using photolithography techniques used every day in micro fabrication. This is beneficial if the electrode would be manufactured on a large scale basis. The technique makes it easy to adapt the array to best fit the structure of interest at the time, in this study the rat cerebellar cortex. The use of SU-8 as the base of the array makes the array very flexible, hence lets it stay close to the same cells following the movement of the brain. The electrodes can then be modified with platinum black to lower the impedance of the electrode up to one order of magnitude, making us able to create smaller electrodes but keeping the low impedance necessary to get a get the signal to noise ratio required. Platinum black modified arrays were also implanted chronically and showed excellent signal recording capabilities in rat cerebellum.

A new type of chip based microelectrode for acute electrophysiological recordings in the CNS has been developed. It's designed to be adaptable to a multitude of specific neuronal environments, in this study the cerebellar cortex of rat and cat. Photolithographically patternened SU-8 is used to yield flexible and biocompatible penetrating shanks with gold leads. Electrodes with an impedance of about 300 kOmega at 1kHz have excellent signal to noise ratio in acute recordings in cat cerebellum.

Neural devices may play an important role in the diagnosis and therapy of several clinical conditions, such as stroke, trauma or neurodegenerative disorders, by facilitating motor and pain control. Such interfaces, chronically implanted in the CNS, need to be biocompatible and have the ability to stimulate and record nerve signals. However, neural devices of today are not fully optimized. Nanostructured surfaces may improve electrical properties and lower evoked tissue responses. Vertical gallium phosphide (GaP) nanowires epitaxially grown from a GaP surface is one way of creating nanostructured electrodes. Thus, we chose to study the soft tissue reactions evoked by GaP surfaces. GaP and the control material titanium (Ti) were implanted in the rat abdominal wall for evaluation of tissue reactions after 1, 6, or 12 weeks. The foreign-body response was evaluated by measuring the reactive capsule thickness and by quantification of ED1-positive macrophages and total cells in the capsule. Furthermore, the concentration of Ga was measured in blood, brain, liver and kidneys. Statistically significant differences were noticed between GaP and Ti at 12 weeks for total and ED1-positive cell densities in the capsule. The chemical analysis showed that the concentration of Ga in brain, liver and kidneys increased during 12 weeks of implantation, indicating loss of Ga from the implant. Taken together, our results show that the biocompatible properties of GaP are worse than those of the well-documented biomaterial Ti.

The electrical properties of the solid state/fluid (Ringer solution) interface for phosphorous- and boron-doped porous silicon are reported and the benefits of using porous silicon as neural recording electrodes are discussed. The impedance, reactance and resistance for doped porous and planar silicon, in Ringer solution, were compared to gold electrodes. Planar silicon displayed approximately a three times higher reactance than porous electrodes. The phosphorous-doped porous electrodes displayed a similar reactance compared to the gold electrodes.

In this study, we developed a microdispenser technique in order to create protein patterns for guidance of neurites from cultured adult mouse dorsal root ganglia (DRG). The microdispenser is a micromachined silicon device that ejects 100 picolitre droplets and has the ability to position the droplets with a precision of 6-8 microm. Laminin and bovine serum albumin (BSA) was used to create adhesive and non-adhesive protein lines on polystyrene surfaces (cell culture dishes). Whole-mounted DRGs were then positioned close to the patterns and neurite outgrowth was monitored. The neurites preferred to grow on laminin lines as compared to the unpatterned plastic. When patterns were made from BSA the neurites preferred to grow in between the lines on the unpatterned plastic surface. We conclude that microdispensing can be used for guidance of sensory neurites. The advantages of microdispensing is that it is fast, flexible, allows deposition of different protein concentrations and enables patterning on delicate surfaces due to its non-contact mode of operation. It is conceivable that microdispensing can be utilized for the creation of protein patterns for guiding neurites to obtain in vitro neural networks, in tissue engineering or rapid screening for guiding proteins.

A micromashined flow-through cell for liquid sampling is presented. The flow channel is formed by two 350 pin deep silicon basins facing each other with a 60 pm diameter orifice in the Centre of one of the basins. A piezo-ceramic disc glued to the silicon basin opposite the orifice controls,the drop ejection in a drop-on-demand fashion. The ejected drops have a diameter of 40 prit, i.e. a drop volume of 34 pl. At an ejection frequency of 100 Hz, a sample flow of 0.2 0/min is obtained.

The main objectives with this investigation are firstly to compare important parameters of a capacitive micromachined ultrasonic transducer (cMUT) for air applications with a corresponding piezoelectric ultrasonic transducer. Secondly, to characterise the cMUT's radiation field using light diffraction tomography. The cMUTs consist of a silicon backplate with small pyramid-shaped cavities and a metallised polymeric membrane. The piezoelectric transducers are nonfocused with a resonance frequency of 450 kHz. Time waveforms, FFT-spectra, S/N-ratio, linearity plots and distance working range are presented for different kinds of combinations of cavity size, membrane thickness and bias voltage. Pressure maps from light diffraction tomography measurements are also presented. The principal conclusions are that the cMUTs have comparable linearity and sensitivity but exhibit superior bandwidth to the compared piezoelectric transducers.

Silicon membranes (diameter=4 mm, thickness=60 /spl mu/m) with a circular matrix of holes centered at the chip were fabricated using anisotropic etching. The perforated area had a diameter of 2 mm. Three membrane geometries (hole sizes of 10, 50 or 100 /spl mu/m) were investigated to find a suitable hole size for the nerve regeneration. The silicon membranes were mounted between two silicone rubber tubes and were implanted in the rat sciatic nerve for 4 or 16 weeks. The function of the regenerated nerve was investigated by measuring the muscle contractility force of the gastrocnemius muscle. The membranes with 100 /spl mu/m holes displayed superior muscle force measurements compared to the chips with 10 /spl mu/m and 50 /spl mu/m holes. The 100 /spl mu/m chip showed a 56% regained muscle force relative the uninjured contralateral side. Cross-sections of the regenerated structures were inspected by light microscopy and SEM. The biocompatibility of the silicon membranes was evaluated by analyzing the tissue reactions 10 days after insertion of the membrane in the rectus muscle of the rat. Titanium discs were used as a reference material. The study showed no obvious difference between titanium and silicon implants.

Two different types of perforated silicon nerve chips with phosphorus doped pick-up electrodes were fabricated. One type of chips held four pick-up electrodes and the other chips were similar to the first but also had a reference electrode and a ground electrode on the chips. The crosstalk between the correct pick-up electrode and neighbouring electrodes was measured with an in vitro set-up, simulating a nerve fibre with a Cu-wire. The second generation of chips which measured the signals of the pick-up electrodes relative to the reference electrode on the chip showed better results than the first second generation of chips displayed an amplitude crosstalk suppression of 20 dB.

The performance of doped silicon and porous silicon as a new electrode material in neurophysiological applications is investigated. By increasing the surface area of adopted silicon region through anodisation in hydrofluoric acid it was found that the impedance parameters of the electrode were considerably improved. Impedance spectra of bare silicon electrodes, porous electrodes and platinum electroplated porous silicon electrodes are presented. It was found that porous silicon electrodes display a 2.5 times higher capacitance and the platinum electroplated porous silicon electrode displayed a 5 times higher capacitance when compared to the corresponding planar doped silicon electrode.

An in vitro model was developed to characterise the electrical properties of silicon microfabricated recording electrodes, using a Cu-wire mimicing a neural signal source. Phosphorous doped electrodes were used to achieve an all silicon device. The model was used to study signal amplitude as a function of distance between the electrode surface and the signal source. Signal crosstalk to neighbouring electrodes on the chips were recorded. The crosstalk was found to be 6 dB using an external reference electrode. Improvements were accomplished with. An on chip reference electrode giving an amplitude crosstalk suppression of 20 dB. It was found that the amplitude decreased by a factor of 2 at a distance of 50 /spl mu/m between the electrode surface and the signal source. Sieve electrodes were also implanted in the rat sciatic nerve and following a 10 week nerve regeneration period the dorsal and ventral (L5) roots in the spinal cord were stimulated. Compound action potentials were recorded via the chip. Lower leg muscle contraction activity was also induced by stimulating the regenerated sciatic nerve via the sieve electrode.

A neural interface could be used to control a limb prosthesis. Such an interface can be achieved by implanting a sieve electrode in the peripheral nerve trunk proximal to an amputation and subsequently record nerve signals intended to control the prostheses. To achieve a long-term stable implantable sieve electrode the implant must fulfil all requirements regarding biocompatibility. This paper therefore addresses cellular reactions in soft tissue induced by silicon and titanium (planar and porous) implants. All implant types displayed the same density of ED1 and ED2 positive macrophages but regarding the capsule formation planar implants displayed an increased capsule formation while the porous implants displayed a constant capsule thickness over time. This paper also addresses the quality of regeneration through four sieve electrode geometries (hole size 30 or 90 /spl mu/m and transparency 20 or 30%). After 12 weeks of nerve regeneration, electrodes with a hole size of 30 /spl mu/m and a 30% transparency displayed the most favourable result as judged by the regained gastrocnemius muscle force.

Square-shaped silicon or titanium implants with plane or porous surfaces surrounded by a rim of silicone were implanted in the rat abdominal wall for evaluation of the tissue response after one, six, or 12 weeks. Cell damage was identified as increased membrane permeability using fluorescence microscopy by injection of propidium iodide prior to the killing of the rats. Capsule thickness and immunohistochemical quantification of macrophages were used as a further measure of the foreign-body reaction. There were no significant differences in capsular cell densities for macrophages, total cells (macrophages, fibroblasts, and other cells), or necrotic cells at the different time points for the four surfaces studied. However, significant differences in the kinetics of the response were found between plane surfaces compared with porous ones. Both types of plane surfaces developed a significant increase in capsule thickness over time in contrast to the porous implants. Porous silicon displayed a significant decrease in total cells in the reactive capsule over time. Furthermore, porous silicon and titanium surfaces displayed a significant decrease in total cell numbers at the implant interface between six and 12 weeks. The present study demonstrated that implanted silicon elicited soft-tissue reactions comparable to that of titanium.

The fabrication process for a silicon micro machined heat exchanger is descibed. A thin walled, 7 μm, channel structure defined by vertical channels, 360 μm deep and 220 μm wide, anisotropically etched in -silicon followed by a pn-etch stop process provided a heat exchanger module in the shape of a “square wave” corrugated membrane. The double sided processing provided vertical channels interdigitated from one side to the other giving a counter flow component with rear side fluid access from the front side. The design allows the stacking of multiple elements to obtain a unit with the desired heat sink capacity.

A flow-through microdispenser is presented. The deadvolume of the dispenser has been reduced to 1.3 μl between the inlet and sampling orifice. The efficiency and stability of the droplet formation has been improved by using a multilayer piezo-ceramic element as actuator. The dispensed droplet volume varies with liquid surface tension and was found to remain stable for concentrations of acetonitrile above 30 % in water. The dispenser was used for sample preparation in MALDI-TOF-MS analysis. Single 110 pl sample droplets containing around 1.8 μM of proteins were analysed. In a pre-concentration experiment a diluted protein sample (around 20 nM) was dispensed in a 4 × 200 droplet sequence for increased sensitivity.