Advanced Micro Devices Corporation

Globalfoundries Inc.

Karsten Wieczorek

Manfred Horstmann

Rolf Stephan

Paul Raymond Besser

Frederick N Hause

Nick Kepler

Thomas Feudel

Thorsten E Kammler

Michael Raab

Larry Lingjer Wang

Stephan Kruegel

Falk Graetsch

Gert Burbach

Kerstin Ruttloff

Martin Trentzsch

Christian Krueger

Patrick Press

Lutz Herrmann

Peter Javorka

Markus Lenski

Matthias Schaller

Christoph Schwan

Christof Streck

Gunter Grasshoff

Scott Luning

Nicholas John Kepler

Larry Yu Wang

Edward E Ehrichs

Wolfgang Buchholtz

Karla Romero

Peter Huebler

Rolf Geilenkeuser

Christian Kr&uuml;ger

Scott Lunning

Christian Radehaus

Stephan Krügel

Bernd Engelmann

Tilo Mantei

Thomas Heller

Mark Craig

Jörg-Oliver Weidner

In a memory cell, the drive current capabilities of the transistors may be adjusted by locally providing an increased gate dielectric thickness and/or gate length of one or more of the transistors of the memory cell. That is, the gate length and/or the gate dielectric thickness may vary along the transistor width direction, thereby providing an efficient mechanism for adjusting the effective drive current capability while at the same time allowing the usage of a simplified geometry of the active region, which may result in enhanced production yield due to enhanced process uniformity. In particular, the probability of creating short circuits caused by nickel silicide portions may be reduced.

Drive current adjustment for transistors by local gate engineering

In a dual stress liner approach, an intermediate etch stop material may be provided on the basis of a plasma-assisted oxidation process rather than by deposition so the corresponding thickness of the etch stop material may be reduced. Consequently, the resulting aspect ratio may be less pronounced compared to conventional strategies, thereby reducing deposition-related irregularities which may translate into a significant reduction of yield loss, in particular for highly scaled semiconductor devices.

Etch stop layer of reduced thickness for patterning a dielectric material in a contact level of closely spaced transistors

By incorporating nitrogen into the P-doped regions and N-doped regions of the gate electrode material prior to patterning the gate electrode structure, yield losses due to reactive wet chemical cleaning processes may be significantly reduced.

Method for reducing defects of gate of CMOS devices during cleaning processes by modifying a parasitic PN junction

By providing a gate dielectric material of increased thickness for P-channel transistors compared to N-channel transistors, degradation mechanisms, such as negative bias threshold voltage instability, hot carrier injection and the like, may be reduced. Due to the enhanced reliability of the P-channel transistors, overall production yield for a specified quality category may be increased, due to the possibility of selecting narrower guard bands for the semiconductor device under consideration.

Gate dielectrics of different thickness in PMOS and NMOS transistors

By forming bulk-like transistors in sensitive RAM areas of otherwise SOI-based CMOS circuits, a significant savings in valuable chip area may be achieved since the RAM areas may be formed on the basis of a bulk transistor configuration, thereby eliminating hysteresis effects that may typically be taken into consideration by providing transistors of increased transistor width or by providing body ties. Hence, the benefit of high switching speed may be maintained in speed-critical circuitry, such as CPU cores, while at the same time the RAM circuit may be formed in a highly space-efficient manner.

Method for manufacturing semiconductor device comprising SOI transistors and bulk transistors

By performing sophisticated anneal techniques, such as laser anneal, flash anneal and the like, for a metal silicide formation, such as nickel silicide, the risk of nickel silicide defects in sensitive device regions, such as SRAM pass gates, may be significantly reduced. Also, the activation of dopants may be performed in a highly localized manner, so that undue damage of gate insulation layers may be avoided when activating and re-crystallizing drain and source regions.

Technique for locally adapting transistor characteristics by using advanced laser/flash anneal techniques

By forming an implantation mask prior to the definition of the drain and the source areas, an effective decoupling of the gate dopant concentration from that of the drain and source concentrations is achieved. Moreover, after removal of the implantation mask, the lateral dimension of the gate electrode may be defined by well-established sidewall spacer techniques, thereby providing a scaling advantage with respect to conventional approaches based on photolithography and anisotropic etching.

Field effect transistor having a doped gate electrode with reduced gate depletion and method of forming the transistor

By predoping of a layer of deposited semiconductor gate material by incorporating dopants during the deposition process, a high uniformity of the dopant distribution may be achieved in the gate electrodes of CMOS devices subsequently formed in the layer of gate material. The improved uniformity of the dopant distribution results in reduced gate depletion and reduced threshold voltage shift in the transistors of the CMOS devices.

CMOS gate structure comprising predoped semiconductor gate material with improved uniformity of dopant distribution and method of forming the structure

An epitaxially grown channel layer is provided on a well structure after ion implantation steps and heat treatment steps are performed to establish a required dopant profile in the well structure. The channel layer may be undoped or slightly doped, as required, so that the finally obtained dopant concentration in the channel layer is significantly reduced compared to a conventional device to thereby provide a retrograde dopant profile in a channel region of a field effect transistor. Additionally, a barrier diffusion layer may be provided between the well structure and the channel layer to reduce up-diffusion during any heat treatments carried out after the formation of the channel layer. The final dopant profile in the channel region may be adjusted by the thickness of the channel layer, the thickness and the composition of the diffusion barrier layer and any additional implantation steps to introduce dopant atoms in the channel layer.

Semiconductor device having a retrograde dopant profile in a channel region

A semiconductor structure comprising a first transistor element and a second transistor element is provided. Stress in channel regions of the first and the second transistor element is controlled by forming stressed layers having a predetermined stress over the transistors. The stressed layers may be used as etch stop layers in the formation of contact vias through an interlayer dielectric formed over the transistors.

Method of forming a semiconductor structure comprising transistor elements with differently stressed channel regions

A method is disclosed in which differing metal layers are sequentially deposited on silicon-containing regions so that the type and thickness of the metal layers may be adapted to specific characteristics of the underlying silicon-containing regions. Subsequently, a heat treatment is performed to convert the metals into metal silicides so as to improve the electrical conductivity of the silicon-containing regions. In this way, silicide portions may be formed that are individually adapted to specific silicon-containing regions so that device performance of individual semiconductor elements or the overall performance of a plurality of semiconductor elements may significantly be improved. Moreover, a semiconductor device is disclosed comprising at least two silicon-containing regions having formed therein differing silicide portions, wherein at least one silicide portion comprises a noble metal.

Method of forming different silicide portions on different silicon-containing regions in a semiconductor device

A method is disclosed in which differing metal layers are sequentially deposited on silicon-containing regions so that the type and thickness of the metal layers may be adapted to specific characteristics of the underlying silicon-containing regions. Subsequently, a heat treatment is performed to convert the metals into metal silicides so as to improve the electrical conductivity of the silicon-containing regions. In this way, silicide portions may be formed that are individually adapted to specific silicon-containing regions so that device performance of individual semiconductor elements or the overall performance of a plurality of semiconductor elements may be significantly improved. Moreover, a semiconductor device is disclosed comprising at least two silicon-containing regions having formed therein differing silicide portions, wherein at least one silicide portion comprises a noble metal.

Semiconductor device having different metal silicide portions and method for fabricating the semiconductor device

By heat treating a silicon dioxide liner prior to patterning a silicon nitride spacer layer, the etch selectivity of the silicon dioxide with respect to the silicon nitride is increased, thereby reducing or eliminating the problem of pitting through the silicon dioxide layer. This allows further scaling of the devices, wherein an extremely thin silicon dioxide liner is required to obtain an accurate lateral patterning of the dopant profile in the drain and source regions.

Method of forming sidewall spacer elements for a circuit element by increasing an etch selectivity

Polysilicon lines are formed, featuring an upper portion extending beyond the lower portion that defines the required CD. Accordingly, metal silicide layers of increased dimensions can be formed on the upper portion of the polysilicon lines so that the resulting gate structures exhibit a very low final sheet resistance. Moreover, in situ sidewall spacers are realized during the process for forming the polysilicon lines and without additional steps and/or costs.

Semiconductor device having T-shaped gate structure comprising in situ sidewall spacers and method of forming the semiconductor device

By maintaining the gate electrode covered during the process flow for forming metal silicide regions in the drain and source of a field effect transistor, an appropriate metal silicide may be formed on the gate electrode which meets the requirement for aggressive gate length scaling. Preferably, a nickel silicide is formed on the gate electrode, whereas the drain and source regions receive the well-established cobalt disilicide. Additionally, the gate electrode dopant profile is effectively decoupled from the drain and source dopant profile.

Growing community of inventors

Dresden, Germany

Karsten Wieczorek