The present disclosure describes a method for forming capping layers configured to prevent the migration of out-diffused cobalt atoms into upper metallization layers In some embodiments, the method includes depositing a cobalt diffusion barrier layer on a liner-free conductive structure that includes ruthenium, where depositing the cobalt diffusion barrier layer includes forming the cobalt diffusion barrier layer self-aligned to the liner-free conductive structure. The method also includes depositing, on the cobalt diffusion barrier layer, a stack with an etch stop layer and dielectric layer, and forming an opening in the stack to expose the cobalt diffusion barrier layer. Finally, the method includes forming a conductive structure on the cobalt diffusion barrier layer.

Metal gate stacks and integrated methods of forming metal gate stacks are disclosed. Some embodiments comprise NbN as a PMOS work function material at a thickness in a range of greater than or equal to 5 Å to less than or equal to 50 Å. The PMOS work function material comprising NbN has an effective work function of greater than or equal to 4.75 eV. Some embodiments comprise HfO.sub.2 as a high-κ metal oxide layer. Some embodiments provide improved PMOS bandedge performance evidenced by improved flatband voltage. Some embodiments exclude transition metal niobium nitride materials as work function materials.

A method of forming a ruthenium film on a substrate by supplying a ruthenium-containing gas includes: forming an adsorption inhibition layer that inhibits adsorption of the ruthenium-containing gas by supplying an adsorption inhibition gas to an end portion and a rear surface of the substrate; transferring the substrate to a chamber; and forming the ruthenium film on the substrate by supplying the ruthenium-containing gas to the chamber.

Semiconductor devices, FinFET devices and methods of forming the same are disclosed. One of the semiconductor devices includes a substrate and a gate strip disposed over the substrate. The gate strip includes a high-k layer disposed over the substrate, an N-type work function metal layer disposed over the high-k layer, and a barrier layer disposed over the N-type work function metal layer. The barrier layer includes at least one first film containing TiAlN, TaAlN or AlN.

According to one or more embodiments, a non-volatile semiconductor memory device includes a semiconductor region, a gate electrode, a charge storage layer, a first insulating layer, a second insulating layers, and a conductive layer. The conductive layer contains titanium (Ti), aluminum (Al) and nitrogen (N) and has a structure in which a plurality of first layers and a plurality of second layers are alternately provided in a thickness direction. Each first layer contains titanium and nitrogen. Each second layer contains aluminum and nitrogen. In the conductive layer, the ratio of aluminum atomic composition to the sum of the titanium atomic composition and the aluminum atomic composition is equal to or less than 50%.

Methods for reducing interface resistance of semiconductor devices leverage dual work function metal silicide. In some embodiments, a method may comprise selectively depositing a metal silicide layer on an Epi surface and adjusting a metal-to-silicon ratio of the metal silicide layer during deposition to alter a work function of the metal silicide layer based on whether the Epi surface is a P type Epi surface or an N type Epi surface to achieve a Schottky barrier height of less than 0.5 eV. The work function for a P type Epi surface may be adjusted to a value of approximately 5.0 eV and the work function for an N type Epi surface may be adjusted to a value of approximately 3.8 eV. The deposition of the metal silicide layer on the Epi surface may be performed prior to deposition of a contact etch stop layer and an activation anneal.

A method includes providing a substrate having a conductive column, a dielectric layer over the conductive column, and a plurality of sacrificial blocks over the dielectric layer, the plurality of sacrificial blocks surrounding the conductive column from a top view; depositing a sacrificial layer covering the plurality of sacrificial blocks, the sacrificial layer having a dip directly above the conductive column; depositing a hard mask layer over the sacrificial layer; removing a portion of the hard mask layer from a bottom of the dip; etching the bottom of the dip using the hard mask layer as an etching mask, thereby exposing a top surface of the conductive column; and forming a conductive material inside the dip, the conductive material being in physical contact with the top surface of the conductive column.

Methods of depositing platinum group metal films of high purity, low resistivity, and good conformality are described. A platinum group metal film is formed in the absence of an oxidant. The platinum group metal film is selectively deposited on a conductive substrate at a temperature less than 200° C. by using an organic platinum group metal precursor.

Semiconductor devices, FinFET devices and methods of forming the same are disclosed. One of the semiconductor devices includes a substrate and a gate strip disposed over the substrate. The gate strip includes a high-k layer disposed over the substrate, an N-type work function metal layer disposed over the high-k layer, and a barrier layer disposed over the N-type work function metal layer. The barrier layer includes at least one first film containing TiAlN, TaAlN or AlN.

A method includes depositing a first conductive layer over a gate dielectric layer; depositing a first work function tuning layer over the first conductive layer; selectively removing the first work function tuning layer from over a first region of the first conductive layer; doping the first work function tuning layer with a dopant; and after doping the first work function tuning layer performing a first treatment process to etch the first region of the first conductive layer and a second region of the first work function tuning layer. The first treatment process etches the first conductive layer at a greater rate than the first work function tuning layer.