Abstract:
An integrated transistor in the form of a nanoscale electromechanical switch eliminates CMOS current leakage and increases switching speed. The nanoscale electromechanical switch features a semiconducting cantilever that extends from a portion of the substrate into a cavity. The cantilever flexes in response to a voltage applied to the transistor gate thus forming a conducting channel underneath the gate. When the device is off, the cantilever returns to its resting position. Such motion of the cantilever breaks the circuit, restoring a void underneath the gate that blocks current flow, thus solving the problem of leakage. Fabrication of the nano-electromechanical switch is compatible with existing CMOS transistor fabrication processes. By doping the cantilever and using a back bias and a metallic cantilever tip, sensitivity of the switch can be further improved. A footprint of the nano-electromechanical switch can be as small as 0.1×0.1 μm2.
Abstract:
A multi-fin FINFET device may include a substrate and a plurality of semiconductor fins extending upwardly from the substrate and being spaced apart along the substrate. Each semiconductor fin may have opposing first and second ends and a medial portion therebetween, and outermost fins of the plurality of semiconductor fins may comprise an epitaxial growth barrier on outside surfaces thereof. The FINFET may further include at least one gate overlying the medial portions of the semiconductor fins, a plurality of raised epitaxial semiconductor source regions between the semiconductor fins adjacent the first ends thereof, and a plurality of raised epitaxial semiconductor drain regions between the semiconductor fins adjacent the second ends thereof.
Abstract:
A strip of sacrificial semiconductor material is formed on top of a non-sacrificial semiconductor material substrate layer. A conformal layer of the non-sacrificial semiconductor material is epitaxially grown to cover the substrate layer and the strip of sacrificial semiconductor material. An etch is performed to selectively remove the strip of sacrificial semiconductor material and leave a hollow channel surrounded by the conformal layer and the substrate layer. Using an anneal, the conformal layer and the substrate layer are reflowed to produce an optical waveguide structure including the hollow channel.
Abstract:
A tunneling transistor is implemented in silicon, using a FinFET device architecture. The tunneling FinFET has a non-planar, vertical, structure that extends out from the surface of a doped drain formed in a silicon substrate. The vertical structure includes a lightly doped fin defined by a subtractive etch process, and a heavily-doped source formed on top of the fin by epitaxial growth. The drain and channel have similar polarity, which is opposite that of the source. A gate abuts the channel region, capacitively controlling current flow through the channel from opposite sides. Source, drain, and gate terminals are all electrically accessible via front side contacts formed after completion of the device. Fabrication of the tunneling FinFET is compatible with conventional CMOS manufacturing processes, including replacement metal gate and self-aligned contact processes. Low-power operation allows the tunneling FinFET to provide a high current density compared with conventional planar devices.
Abstract:
A dual width SOI FinFET is disclosed in which different portions of a strained fin have different widths. A method of fabrication of such a dual width FinFET entails laterally recessing the strained fin in the source and drain regions using a wet chemical etching process so as to maintain a high degree of strain in the fin while trimming the widths of fin portions in the source and drain regions to less than 5 nm. The resulting FinFET features a wide portion of the fin in the channel region underneath the gate, and a narrower portion of the fin in the source and drain regions. An advantage of the narrower fin is that it can be more easily doped during the growth of the epitaxial raised source and drain regions.
Abstract:
A dual width SOI FinFET is disclosed in which different portions of a strained fin have different widths. A method of fabrication of such a dual width FinFET entails laterally recessing the strained fin in the source and drain regions using a wet chemical etching process so as to maintain a high degree of strain in the fin while trimming the widths of fin portions in the source and drain regions to less than 5 nm. The resulting FinFET features a wide portion of the fin in the channel region underneath the gate, and a narrower portion of the fin in the source and drain regions. An advantage of the narrower fin is that it can be more easily doped during the growth of the epitaxial raised source and drain regions.
Abstract:
An integrated circuit die includes a substrate having a first layer of semiconductor material, a layer of dielectric material on the first layer of semiconductor material, and a second layer of semiconductor material on the layer of dielectric material. An extended channel region of a transistor is positioned in the second layer of semiconductor material, interacting with a top surface, side surfaces, and potentially portions of a bottom surface of the second layer of semiconductor material. A gate dielectric is positioned on a top surface and on the exposed side surface of the second layer of semiconductor material. A gate electrode is positioned on the top surface and the exposed side surface of the second layer of semiconductor material.
Abstract:
A resistive random access memory (RRAM) structure is formed on a supporting substrate and includes a first electrode and a second electrode. The first electrode is made of a silicided fin on the supporting substrate and a first metal liner layer covering the silicided fin. A layer of dielectric material having a configurable resistive property covers at least a portion of the first metal liner. The second electrode is made of a second metal liner layer covering the layer of dielectric material and a metal fill in contact with the second metal liner layer. A non-volatile memory cell includes the RRAM structure electrically connected between an access transistor and a bit line.
Abstract:
A method for forming a complementary metal oxide semiconductor (CMOS) semiconductor device includes providing a stressed silicon-on-insulator (sSOI) wafer comprising a stressed semiconductor layer having first and second laterally adjacent stressed semiconductor portions. The first stressed semiconductor portion defines a first active region. The second stressed semiconductor portion is replaced with an unstressed semiconductor portion. The unstressed semiconductor portion includes a first semiconductor material. The method further includes driving a second semiconductor material into the first semiconductor material of the unstressed semiconductor portion defining a second active region.
Abstract:
A strip of sacrificial semiconductor material is formed on top of a non-sacrificial semiconductor material substrate layer. A conformal layer of the non-sacrificial semiconductor material is epitaxially grown to cover the substrate layer and the strip of sacrificial semiconductor material. An etch is performed to selectively remove the strip of sacrificial semiconductor material and leave a hollow channel surrounded by the conformal layer and the substrate layer. Using an anneal, the conformal layer and the substrate layer are reflowed to produce an optical waveguide structure including the hollow channel.