Limits...
Improved selectivity from a wavelength addressable device for wireless stimulation of neural tissue.

Seymour EÇ, Freedman DS, Gökkavas M, Ozbay E, Sahin M, Unlü MS - Front Neuroeng (2014)

Bottom Line: Optical activation provides a wireless means of energy transfer to the neurostimulator, eliminating wires and the associated complications.We assessed the improved addressability of individual devices via wavelength selectivity as compared to spatial selectivity alone through on-bench optical measurements of the devices in combination with an in vivo light intensity profile in the rat cortex obtained in a previous study.We show that wavelength selectivity improves the individual addressability of the floating stimulators, thus increasing the number of devices that can be implanted in close proximity to each other.

View Article: PubMed Central - PubMed

Affiliation: Department of Biomedical Engineering, Boston University, Boston MA, USA.

ABSTRACT
Electrical neural stimulation with micro electrodes is a promising technique for restoring lost functions in the central nervous system as a result of injury or disease. One of the problems related to current neural stimulators is the tissue response due to the connecting wires and the presence of a rigid electrode inside soft neural tissue. We have developed a novel, optically activated, microscale photovoltaic neurostimulator based on a custom layered compound semiconductor heterostructure that is both wireless and has a comparatively small volume (<0.01 mm(3)). Optical activation provides a wireless means of energy transfer to the neurostimulator, eliminating wires and the associated complications. This neurostimulator was shown to evoke action potentials and a functional motor response in the rat spinal cord. In this work, we extend our design to include wavelength selectivity and thus allowing independent activation of devices. As a proof of concept, we fabricated two different microscale devices with different spectral responsivities in the near-infrared region. We assessed the improved addressability of individual devices via wavelength selectivity as compared to spatial selectivity alone through on-bench optical measurements of the devices in combination with an in vivo light intensity profile in the rat cortex obtained in a previous study. We show that wavelength selectivity improves the individual addressability of the floating stimulators, thus increasing the number of devices that can be implanted in close proximity to each other.

No MeSH data available.


Related in: MedlinePlus

Fabrication steps of GaAs FLAME stimulators (A) GaAs/AlGaAs wafer structure, (B) etching down to the p-layer, (C) etching down to n-layer, (D) p metal deposition: Ti:Au, (E) n-metal deposition: Ge:Au:Ge:Au:Ni:Au, (F) silicon nitride (SiN) deposition, (G) etching SiN from the contacts, (H) top release: chips are diced 150 μm-deep around the devices, (I) backside etching to complete the release of devices, (J) the micrograph of a completed FLAME device with the arrow indicating the cross section shown in (A–I).
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC3927122&req=5

Figure 3: Fabrication steps of GaAs FLAME stimulators (A) GaAs/AlGaAs wafer structure, (B) etching down to the p-layer, (C) etching down to n-layer, (D) p metal deposition: Ti:Au, (E) n-metal deposition: Ge:Au:Ge:Au:Ni:Au, (F) silicon nitride (SiN) deposition, (G) etching SiN from the contacts, (H) top release: chips are diced 150 μm-deep around the devices, (I) backside etching to complete the release of devices, (J) the micrograph of a completed FLAME device with the arrow indicating the cross section shown in (A–I).

Mentions: Two-inch diameter wafers were diced into 9 mm × 10 mm pieces with the DISCO automatic dicer and the layer structures were verified with scanning electron microscopy (Zeiss). Fabrication consisted of standard positive photolithography, wet and dry etching, metallization and passivation steps, which are shown in Figure 3. Mask layouts were drawn using the Virtuoso tool (Cadence). Mask designs were transferred to iron oxide masks with the DWL 66 mask writer (Heidelberg Instruments). Series photodiodes were implemented by making the p-contact on the p-type layer of the first (top) photodiode and the n-contact on the n-type layer of the second (bottom) photodiode. Two series photodiodes were connected to each other through a highly doped tunneling junction. The solution for etching to the p-type layer was prepared by mixing 25:5:1 (volume):deionized (DI) H2O:phosphoric acid:H2O2. The p-type metal ohmic contact was made by evaporating Pt:Ti:Pt:Au (100:400:100:1500 Å) metallization layers (bottom to top) using electron beam evaporation. The P-metal was subsequently annealed at 400°C for 1 min. We utilized a thin (10 nm) layer of AlAs that was grown on top of the bottom n-layer as an etch stop. A selective etchant was used to etch to the n-layer and was prepared by dissolving 1 g of citric acid monohydrate crystals in 2 ml of water. This citric acid solution (CAS) was mixed with hydrogen peroxide: 10 CAS:1 H2O2. This solution selectively etches GaAs as compared to AlGaAs layers (Kim, 1998). The chips were dipped into a solution of 1 BOE (buffered oxide etch):15 DI H2O for 15 s to remove the native oxide layer before etching. The AlAs layer is removed by dipping the sample into a solution of 1 BOE:15 DI H2O for 15 s. The n-type contact was fabricated by evaporating Ge:Au:Ge:Au:Ni:Au (60:100:100:240:100:1500 Å) metallization layers (bottom to top) and annealed at 430°C for 1 min. After contacts were made the devices were tested under a probe station to evaluate their electrical characteristics. An insulating layer of 360 nm of silicon nitride was deposited using plasma enhanced chemical vapor deposition (PECVD). Contacts were exposed by etching silicon nitride with reactive ion etching (RIE) using 50 sccm SF6 at 100 mTorr pressure and 150 W power. Individual devices were removed from the chips by a release procedure that consisted of creating 150 μm cuts around the devices by the DISCO automatic dicer and then backside etching with 4:1:citric acid (1 M):H2O2 (30%) at 50°C until devices separated from the chip die. Figure 4 shows a micrograph of a fabricated shank device and the relative size of a typical FLAME device.


Improved selectivity from a wavelength addressable device for wireless stimulation of neural tissue.

Seymour EÇ, Freedman DS, Gökkavas M, Ozbay E, Sahin M, Unlü MS - Front Neuroeng (2014)

Fabrication steps of GaAs FLAME stimulators (A) GaAs/AlGaAs wafer structure, (B) etching down to the p-layer, (C) etching down to n-layer, (D) p metal deposition: Ti:Au, (E) n-metal deposition: Ge:Au:Ge:Au:Ni:Au, (F) silicon nitride (SiN) deposition, (G) etching SiN from the contacts, (H) top release: chips are diced 150 μm-deep around the devices, (I) backside etching to complete the release of devices, (J) the micrograph of a completed FLAME device with the arrow indicating the cross section shown in (A–I).
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC3927122&req=5

Figure 3: Fabrication steps of GaAs FLAME stimulators (A) GaAs/AlGaAs wafer structure, (B) etching down to the p-layer, (C) etching down to n-layer, (D) p metal deposition: Ti:Au, (E) n-metal deposition: Ge:Au:Ge:Au:Ni:Au, (F) silicon nitride (SiN) deposition, (G) etching SiN from the contacts, (H) top release: chips are diced 150 μm-deep around the devices, (I) backside etching to complete the release of devices, (J) the micrograph of a completed FLAME device with the arrow indicating the cross section shown in (A–I).
Mentions: Two-inch diameter wafers were diced into 9 mm × 10 mm pieces with the DISCO automatic dicer and the layer structures were verified with scanning electron microscopy (Zeiss). Fabrication consisted of standard positive photolithography, wet and dry etching, metallization and passivation steps, which are shown in Figure 3. Mask layouts were drawn using the Virtuoso tool (Cadence). Mask designs were transferred to iron oxide masks with the DWL 66 mask writer (Heidelberg Instruments). Series photodiodes were implemented by making the p-contact on the p-type layer of the first (top) photodiode and the n-contact on the n-type layer of the second (bottom) photodiode. Two series photodiodes were connected to each other through a highly doped tunneling junction. The solution for etching to the p-type layer was prepared by mixing 25:5:1 (volume):deionized (DI) H2O:phosphoric acid:H2O2. The p-type metal ohmic contact was made by evaporating Pt:Ti:Pt:Au (100:400:100:1500 Å) metallization layers (bottom to top) using electron beam evaporation. The P-metal was subsequently annealed at 400°C for 1 min. We utilized a thin (10 nm) layer of AlAs that was grown on top of the bottom n-layer as an etch stop. A selective etchant was used to etch to the n-layer and was prepared by dissolving 1 g of citric acid monohydrate crystals in 2 ml of water. This citric acid solution (CAS) was mixed with hydrogen peroxide: 10 CAS:1 H2O2. This solution selectively etches GaAs as compared to AlGaAs layers (Kim, 1998). The chips were dipped into a solution of 1 BOE (buffered oxide etch):15 DI H2O for 15 s to remove the native oxide layer before etching. The AlAs layer is removed by dipping the sample into a solution of 1 BOE:15 DI H2O for 15 s. The n-type contact was fabricated by evaporating Ge:Au:Ge:Au:Ni:Au (60:100:100:240:100:1500 Å) metallization layers (bottom to top) and annealed at 430°C for 1 min. After contacts were made the devices were tested under a probe station to evaluate their electrical characteristics. An insulating layer of 360 nm of silicon nitride was deposited using plasma enhanced chemical vapor deposition (PECVD). Contacts were exposed by etching silicon nitride with reactive ion etching (RIE) using 50 sccm SF6 at 100 mTorr pressure and 150 W power. Individual devices were removed from the chips by a release procedure that consisted of creating 150 μm cuts around the devices by the DISCO automatic dicer and then backside etching with 4:1:citric acid (1 M):H2O2 (30%) at 50°C until devices separated from the chip die. Figure 4 shows a micrograph of a fabricated shank device and the relative size of a typical FLAME device.

Bottom Line: Optical activation provides a wireless means of energy transfer to the neurostimulator, eliminating wires and the associated complications.We assessed the improved addressability of individual devices via wavelength selectivity as compared to spatial selectivity alone through on-bench optical measurements of the devices in combination with an in vivo light intensity profile in the rat cortex obtained in a previous study.We show that wavelength selectivity improves the individual addressability of the floating stimulators, thus increasing the number of devices that can be implanted in close proximity to each other.

View Article: PubMed Central - PubMed

Affiliation: Department of Biomedical Engineering, Boston University, Boston MA, USA.

ABSTRACT
Electrical neural stimulation with micro electrodes is a promising technique for restoring lost functions in the central nervous system as a result of injury or disease. One of the problems related to current neural stimulators is the tissue response due to the connecting wires and the presence of a rigid electrode inside soft neural tissue. We have developed a novel, optically activated, microscale photovoltaic neurostimulator based on a custom layered compound semiconductor heterostructure that is both wireless and has a comparatively small volume (<0.01 mm(3)). Optical activation provides a wireless means of energy transfer to the neurostimulator, eliminating wires and the associated complications. This neurostimulator was shown to evoke action potentials and a functional motor response in the rat spinal cord. In this work, we extend our design to include wavelength selectivity and thus allowing independent activation of devices. As a proof of concept, we fabricated two different microscale devices with different spectral responsivities in the near-infrared region. We assessed the improved addressability of individual devices via wavelength selectivity as compared to spatial selectivity alone through on-bench optical measurements of the devices in combination with an in vivo light intensity profile in the rat cortex obtained in a previous study. We show that wavelength selectivity improves the individual addressability of the floating stimulators, thus increasing the number of devices that can be implanted in close proximity to each other.

No MeSH data available.


Related in: MedlinePlus