Christopher Locke

Single-crystal cubic silicon carbide: An in vivo biocompatible semiconductor for brain machine interface devices

Single crystal silicon carbide (SiC) is a wide band-gap semiconductor which has shown both bio- and hemo-compatibility [1–5]. Although single crystalline SiC has appealing bio-sensing potential, the material has not been extensively characterized. Cubic silicon carbide (3C-SiC) has superior in vitro biocompatibility compared to its hexagonal counterparts [3, 5]. Brain machine interface (BMI) systems using implantable neuronal prosthetics offer the possibility of bi-directional signaling, which allow sensory feedback and closed loop control. Existing implantable neural interfaces have limited long-term reliability, and 3C-SiC may be a material that may improve that reliability. In the present study, we investigated in vivo 3C-SiC biocompatibility in the CNS of C56BL/6 mice. 3C-SiC was compared against the known immunoreactive response of silicon (Si) at 5, 10, and 35 days. The material was examined to detect CD45, a protein tyrosine phosphatase (PTP) expressed by activated microglia and macrophages. The 3C-SiC surface revealed limited immunoresponse and significantly reduced microglia compared to Si substrate.

3C-SiC Films on Si for MEMS Applications: Mechanical Properties

Single crystal 3C-SiC films were grown on (100) and (111) Si substrate orientations in order to study the resulting mechanical properties of this material. In addition, poly-crystalline 3C-SiC was also grown on (100)Si so that a comparison with monocrystaline 3C-SiC, also grown on (100)Si, could be made. The mechanical properties of single crystal and polycrystalline 3C-SiC films grown on Si substrates were measured by means of nanoindentation using a Berkovich diamond tip. These results indicate that polycrystalline SiC thin films are attractive for MEMS applications when compared with the single crystal 3C-SiC, which is promising since growing single crystal 3C-SiC films is more challenging. MEMS cantilevers and membranes fabricated from a 2 µm thick single crystal 3C-SiC grown on (100)Si under similar conditions resulted in a small degree of bow with only 9 µm of deflection for a cantilever of 700 µm length with an estimated tensile film stress of 300 MPa. Single crystal 3C-SiC films on (111)Si substrates have the highest elastic and plastic properties, although due to high residual stress they tend to crack and delaminate.

Atomic layer deposition of Pb(Zr,Ti)Ox on 4H-SiC for metal-ferroelectric-insulator-semiconductor diodes

Atomic layer deposited (ALD) Pb(Zr,Ti)Ox (PZT) ultra-thin films were synthesized on an ALD Al2O3 insulation layer on 4H-SiC substrate for metal-ferroelectric-insulator-semiconductor (MFIS) device applications. The as-deposited PZT was amorphous but crystallized into a perovskite polycrystalline structure with a preferred [002] orientation upon rapid thermal annealing (RTA) at 950 °C. The capacitance-voltage and current-voltage characteristics of the MFIS devices indicate carrier injection to the film induced by polarization and Fowler-Nordheim (FN) tunneling when electric field was high. The polarization-voltage measurements exhibited reasonable remanent and saturation polarization and a coercive electrical field comparable to that reported for bulk PZT. The piezoresponse force microscope measurements confirmed the polarization, coercive, and retention properties of ultra-thin ALD PZT films.

Low Stress Heteroepitaxial 3C-SiC Films Characterized by Microstructure Fabrication and Finite Elements Analysis

Chemical vapor deposition in the low pressure regime of a high quality 3C-SiC film on silicon (100)-oriented substrates was carried out using silane (SiH 4 ), propane (C 3 H 8 ), and hydrogen (H 2 ) as the silicon supply, carbon supply, and gas carrier, respectively. The resulting bow in the freestanding cantilever structures was evaluated by an optical profilometer, and the residual gradient stress (σ 1 ) in the films was calculated to be approximately between 15 and 20 MPa, which is significantly lower than the previously reported 3C-SiC on Si films. Finite element simulations of the stress field in the cantilever have been carried out to separate the uniform contribution (σ 0 ), related to the SiC/Si interface, from the gradient one (σ 1 ), related to the defects present in the SiC epilayer.

High Quality Single Crystal 3C-SiC(111) Films Grown on Si(111)

We have developed a high-quality growth process for 3C-SiC on on-axis (111)Si substrates with the ultimate goal to demonstrate high quality and yield electronic and MEMS devices. A single-side polished 50 mm (111)Si wafer was loaded into a hot-wall SiC CVD reactor for growth. The 3C-SiC process was performed in two stages: carbonization in propane and hydrogen at 1135°C and 400 Torr followed by growth at 1380°C and 100 Torr. X-ray diffraction rocking curve analysis of the 3C-SiC(222) peak indicates a FWHM value of 219 arcsec. This is a very interesting result given that the film thickness was only 2 µm, thus indicating that the grown film is of very high quality compared with published literature values. X-ray polar figure mapping was performed and it was observed that the micro twin content was below the detection limit. Therefore TEM characterization was performed in plan view to allow assessment of the stacking fault density as well as confirmation of the very low micro twin concentration in this film. TEM analysis indicates a low concentration of stacking faults in the range of 104 cm-1.

3C-SiC Film Growth on Si Substrates

The aim of this work is to give an overview on 3C-SiC growth on Si substrates. Starting from the reasons why SiC is considered such an interesting innovative material, with a survey of application already demonstrated, we will present data explaining the most important issues in this hetero-epitaxy system and how the chemical vapor deposition process influences the resulting 3C-SiC film properties. 3C-SiC crystal structure is strongly dependent on the process parameters within the reaction chamber during growth as well as the substrate surface properties. Part of this work is then focused on the main crystallographic defects characterizing the 3CSiC/Si system and on the resulting wafer bow due to the large misfit between the materials. Defects and wafer bow, are a direct consequence of the large stress generated at the interface. The work closes discussing the encouraging improvements in 3C-SiC crystal quality obtained by the introduction of compliant Si substrates.

Growth Rate Effect on 3C-SiC Film Residual Stress on (100) Si Substrates

SiC is a candidate material for micro- and nano-electromechanical systems (MEMS and NEMS). In order to understand the impact that the growth rate has on the residual stress of CVD-grown 3C-SiC hetero-epitaxial films on Si substrates, growth experiments were performed and the resulting stress was evaluated. Film growth was performed using a two-step growth process with propane and silane as the C and Si precursors in hydrogen carrier gas. The film thickness was held constant at ~2.5 µm independent of the growth rate so as to allow for direct films comparison as a function of the growth rate. Supported by profilometry, Raman and XRD analysis, this study shows that the growth rate is a fundamental parameter for low-defect and low-stress hetero-epitaxial growth process of 3C-SiC on Si substrates. XRD (rocking curve analysis) and Raman spectroscopy show that the crystal quality of the films increases with decreasing growth rate. From curvature measurements, the average residual stress within the layer using the modified Stoney’s equation was calculated. The results show that the films are under compressive stress and the calculated residual stress also increases with growth rate, from -0.78 GPa to -1.11 GPa for 3C-SiC films grown at 2.45 and 4 µm/h, respectively.

Silicon carbide neural implants: In vivo neural tissue reaction

Novel materials are needed for intracortical neural implants (INI) to extend their reliability and functionality beyond a few years. Cubic silicon carbide (3C-SiC) is a chemically inert, physically robust semiconductor that has shown, through extensive in vitro testing, a high biocompatibility with neural cells. Recently we have shown that 3C-SiC does not attract a negative immune response from microglia in vivo, but the implants size did not allow adequate investigation of tissue response [1]. We produced a passive implant to test the in vivo tissue reaction of C57BL/6J mice to 3C-SiC and compare to our positive control of silicon (Si). Dual, triangular shanks were fabricated from each material and combined into a single device which was then implanted simultaneously into three C57BL/6J mouse brains for 35 days. The mice were perfused with 4% paraformaldehyde and the brains treated using immunohistochemistry. Fluorescence microscopy indicated that Si produced the expected increased inflammatory response from both microglia/ macrophage and astrocyte cells, whereas 3C-SiC shows minimal inflammatory reaction from these glial cells. Si also created tissue voids larger than the implants themselves whereas 3C-SiC showed minimal voids and even still had neuronal processes in contact with the implant. Our conclusion is that 3C-SiC shows great potential for use within the neural environment and should be fashioned into active INI to evaluate signal quality over time.

The Development of Silicon Carbide Based Electrode Devices for Central Nervous System Biomedical Implants

Brain machine interface (BMI) technology has been demonstrated to be a therapeutic solution for assisting people suffering from damage to the central nervous system (CNS), but BMI devices using implantable neural prosthetics have experienced difficulties in that they are recognized by glial cells as being foreign material, which leads to an immune response cascade process called gliosis. One material, cubic silicon carbide (3C-SiC), may provide an excellent solution for the generation of an implantable neural prosthetic interface component of a BMI system. We have recently reported on the biocompatibility of 3C-SiC with immortalized cells, and have extended this work by demonstrating neural cell action potential instigation via an electrode type device. Biocompatibility assessment of 3C-SiC was accomplished using in vitro methodology. 96 hour MTT [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide] assays were performed to determine neural cell viability. Atomic force microscopy (AFM) was used to quantify attached cell morphology and determine lamellipodia/ filopodia interaction with the surface of the semiconductor. It was seen that neurons show excellent viability, cell morphology, and good lamellipodia/ filopodia permissiveness when interacting with 3C-SiC. A neuronal activation device (NAD), based on the planar Michigan microelectrode probe, was constructed from 3C-SiC with the goal of activating an action potential within a neuron. In order to illicit an action potential, neurons were seeded on the NAD device and then they were subjected to a biphasic square pulse signal. Successful action potential activation was recorded through the use of Rhod-2, a Ca 2+ sensitive fluorescent dye. Based on these results, 3C-SiC may be an excellent material platform for neural prosthetics.

Cellular Interactions on Epitaxial Graphene on SiC (0001) Substrates

An ever-increasing demand for biocompatible materials provides motivation for the development of advanced materials for challenging applications ranging from disease detection to organ function restoration. Carbon-based materials are considered promising candidates because they combine good biocompatibility with high chemical resistance. In this work we present an initial assessment of the biocompatibility of epitaxial graphene on 6H-SiC(0001). We have analyzed the interaction of HaCaT (human keratinocyte) cells on epitaxial graphene and compared it with that on bare 6H-SiC(0001). We have found that for both graphene and 6H-SiC there is evidence of cell-cell and cell substrate interaction which is normally an indication of the biocompatibility of the material.