Search

With the emergence of 3D integration and wafer level packaging, the pillar bumping process has become a critical processing step. As the process has matured, significant efforts have been made for optimization in terms of both production and cost. This is especially true for photoresist stripping, since thick photoresists are needed to pattern high aspect ratios and must be removed in subsequent steps to expose underlying metal films. In this study, the removal of a thick negative tone photoresist with solder pillar plated wafers were investigated using a dimethyl sulfoxide blend with a quaternary ammonium hydroxide. Emphasis was placed on exploring the feasibility of an extended bath life beyond 7 days. Coupon level beaker tests were used to provide insight into the swelling and dissolution mechanisms with fresh and aged chemistry. Additionally, several chemistries were compared to examine the effect of the quaternary ammonium hydroxide. Based on these findings, chemical aging studies were performed on 300 mm wafers using a highly customizable single wafer processing tool that combined soaking and a high-pressure spray processing to potentially prolong the bath age even further. In addition to stripping efficiency at extended bath life, the compatibly with pillar structures and seed layers is investigated. Results from the study highlight potential methods to extend bath life while maintaining photoresist stripping efficiency.

We demonstrate a novel process for building a Resistive RAM (ReRAM) stack which reduces the forming voltage ( $\text{V}_{\textit {form}}$ ) and increases the switching resistance, both characteristics that are important ingredients for the use of ReRAM in scalable analog compute for AI. Utilizing this process, we explore analog switching characteristics above 100k $\Omega $ and demonstrate 4-bit programming at Rmax $=1\text{M}\Omega $ . Utilizing the same writing characteristics, CIFAR-10 inference simulation shows 90% accuracy, comparable to the full precision model accuracy.

An analysis of NanoSheet (NS) transistor parasitic resistance components is presented and correlated to the resistance readout on Si wafers. With this model, it is possible to identify which components cause the parasitic resistance increases as CPP (contacted poly-Si pitch) further scales from 48 nm to 44 nm pitch. In this study, an alternative MOL (middle of line) metallization scheme is implemented to reduce circuit RC delay. This model helps to further reduce the transistor parasitic resistance from NFET/PFET S/D (source/drain) epitaxy or silicide. As CPP scales, NS PFET parasitic resistance reduction is more challenging and requires optimization in S/D epitaxy, silicide and metallization. Based on parasitic resistance modeling, we present a new wrap-around contact structure which eliminates the vertical epi resistance component, hence reducing overall device resistance.

We report that n-dipole and p-dipole (dual dipoles) can be co-integrated to provide a more flexible volumeless multiple threshold voltage(multi-Vt) solution in FinFET and Nanosheet (NS) technologies. The p-dipole process for dual dipoles co-integration is identified. When the Vt shift is less than 100m V, the mobility is slightly degraded, but other properties are not clearly affected. The improved pFET performance is from the Vt reduction. The dipole co-integration also provides a novel method for Vt definition via dipole Vt compensation. Our selective dipole enablement can implement near bandedge (BE) multi- Vt for high performance application.

In this paper, full bottom dielectric isolation (BDI) is first demonstrated on horizontally stacked Nanosheet device structures with Lmetal 12 nm. The comparison of full BDI scheme vs punch through stopper (PTS) scheme has been systematically studied. By comparing off-state leakage current, short channel behavior and effective capacitance (Ceff) for both schemes, we show that BDI could potentially provide: 1) good immunity of sub-channel leakage due to process variation (from parasitic "fat-Fin" which is unique in Nanosheet structure); 2) power-performance co-optimization.

In Nanosheet (NS) device architecture, it is much more challenging than FinFET to develop a suitable multiple threshold voltage (multi-Vt) integration with more restrictive requirement on the dimensions due to the critical dimension scaling and complex structure. In this abstract, we reported an innovative integration scheme to enable volumeless multi-Vt and metal multi-Vt to provide the multi-Vt solutions in NS technology for high performance computing (HPC) and low-power applications. We developed a new volumeless multi-Vt for NS to solve the device geometry constraint and offer more margin and the opportunity for further sheet-to-sheet spacing (Tsus) reduction. Furthermore, metal gate boundary control (MGBC) was developed to enable variable NS widths on the same wafer to satisfy both HPC and low-power applications.

In this paper, we present for the first time a “Gate-Cut-Last” integration scheme completed within the Replacement Metal Gate (RMG) module. This novel gate cut (CT) technique allows the scaling of gate extension length past the end fin which reduces parasitic capacitance, leakage and performance variation. In addition, we demonstrate that CT-in-RMG is a promising alternative integration process that can enable scaling for future logic technology nodes. Device, circuit and reliability results are shown to compare this novel CT-in-RMG process to the conventional gate cut method.

In this study, a manufacturable CMOS dual solid phase epitaxy (SPE) process with pc < 2.2×10−9 Q-cm2 on both NFET and PFET is demonstrated on the hardware with 7nm ground rule. Contact resistivity reduction strategies of both the conventional approach of high in-situ doped epi and the novel SPE processes are systematically studied on device and ring oscillator (RO) level. Clear improvement in the RO delay is accomplished by the novel dual SPE process on the CMOS flow. Stronger performance benefit is demonstrated with smaller contact sizes towards future CMOS technology nodes.

In this paper, for the first time we demonstrate that horizontally stacked gate-all-around (GAA) Nanosheet structure is a good candidate for the replacement of FinFET at the 5nm technology node and beyond. It offers increased W eff per active footprint and better performance compared to FinFET, and with a less complex patterning strategy, leveraging EUV lithography. Good electrostatics are reported at L g =12nm and aggressive 44/48nm CPP (Contacted Poly Pitch) ground rules. We demonstrate work function metal (WFM) replacement and multiple threshold voltages, compatible with aggressive sheet to sheet spacing for wide stacked sheets. Stiction of sheets in long-channel devices is eliminated. Dielectric isolation is shown on standard bulk substrate for sub-sheet leakage control. Wrap-around contact (WAC) is evaluated for extrinsic resistance reduction.

We present a 7nm technology with the tightest contacted poly pitch (CPP) of 44/48nm and metallization pitch of 36nm ever reported in FinFET technology. To overcome optical lithography limits, Extreme Ultraviolet Lithography (EUV) has been introduced for multiple critical levels for the first time. Dual strained channels have been also implemented to enhance mobility for high performance applications.

In the attempts to push the resolution limits of 193nm immersion lithography, this work demonstrates the building of 3 metal level 56nm pitch copper dual-damascene interconnects, using Negative-Tone Development Lithography-Etch-Lithography-Etch (LELE) Patterning at line level. Line Resistance and intra-level capacitance can be affected by the double patterning integration, but a good process window has been demonstrated, showing no impact on RC performance. The introduction of a self-aligned via (SAV) process with a TiN hard-mask is able to provide a robust process window in terms of via-metal short yield at line and via level. SAV implementation at these dimensions also affects the aspect-ratio of the structures and leads to new challenges in metallization: optimized profile, without bowing or undercut, is mandatory to enable the filling of 28nm lines. The metal hard-mask has to be removed or at least faceted by erosion. This can be achieved by conventional RIE process optimization, but pushes the RIE selectivity to challenging limits. Physical dimensions on target and via chain yield have been demonstrated by fine tuning RIE process. Profile improvement can also be achieved by the introduction of new WET process, helping the removal of the metal hard-mask while being neutral to the ULK. We have demonstrated good yield and reliability with an integration using hard-mask wet removal.

SiGe for channel material has been explored as a major technology element after the introduction of FINFET into CMOS technology [1–4]. Research on long channel FETs and discrete short channel FETs demonstrated benefits in mobility [1–4] and reliability [2]. Given the disruption that SiGe FIN brings, every aspect associated with SiGe FIN needs to be carefully studied towards technology insertion. In this paper, we report the latest SiGe-based FINFET CMOS technology development. CMOS FINFETs with Si-FIN nFET and SiGe-FIN pFET is demonstrated as a viable technology solution for both server and mobile applications at 10nm node and beyond.

Stacked horizontal Gate All Around Nanosheet transistors were recently proposed as a replacement of FinFET for the sub-7nm device nodes. In order to stack Nanosheet channels, a specific SiGe/Si superlattice epitaxy is employed, with SiGe being used as a sacrificial layer that will define the suspension thickness (Tsus) between the channels. In this paper, we will consider the optimization of SiGe/Si multilayer epitaxy for Nanosheet channel definition (Fig. 1). Stacks with different Ge compositions are physically characterized and studied. With strain characterization and x-ray diffraction measurements on (100) substrates, we will show that it is possible to form pseudomorphic defect-free SiGe/Si multilayers over a wide range of Ge concentrations in the sacrificial SiGe (Fig. 2 and 3). The uniformity obtained on blanket 300mm substrates will be presented (Fig. 4), with a discussion on the SiGe/Si interface abruptness. In the context of CMOS device integration, it is necessary to employ a low temperature STI process to mitigate the Ge inter-diffusion in the silicon channel and to avoid the possible generation of defects from plastic relaxation of SiGe. The SiGe and Ge-Si intermixed regions are removed with appropriate selective etch process in two Nanosheet specific modules: inner spacer formation in the extension regions (Fig. 5) and post channel release in the gate regions (Fig. 6), leaving only 5nm intrinsic silicon Nanosheets. We will report the etch rate and selectivity of our in-house vapor-phase channel release process performed at low temperature. This process enables us to remove SiGe versus Si with a great selectivity and a limited amount of silicon loss during channel release, before gate stack deposition around the stacked Nanosheets. From a device perspective, we will show that the optimization of the silicon thickness in the SiGe/Si superlattice plays an important role in the Nanosheet device performance and electrostatics (Fig. 7). Finally, with a good SiGe/Si multilayer uniformity control, a reduced thermal budget and after optimization of our in-house SiGe selective etch process, we will show the capability to form 5nm silicon Nanosheet channels with small within wafer variation, low device variability and good process control, compatible with the fabrication of very large scale integrated circuits for future CMOS device nodes. Acknowledgements: This work was performed by the Research Alliance Teams at various IBM Research Facilities. Figure 1

With transistor scaling in 7nm technology and beyond, sacrificial silicon materials etches (e.g. dummy poly silicon gate removal) are considered to be among the most challenging hurdle in FinFET process development. In this paper, we present a dry chemical etch technique for selective etching of single crystal, poly-crystal and amorphous silicon on various FinFET device process steps. It was demonstrated that this technique could completely remove poly silicon in vertically high aspect ratio (AR>5) FinFET gates with a large process window (over-etch budget ~200%) while achieving the lowest gate leakage current and best short channel FET yield. Proper surface preparation, queue time control and etch by-product removal strategies are discussed. The residue free etch and etch by-product sublimation mechanisms are also investigated by High Resolution Electron Microscopy (HREM) and Fourier Transform Infrared Spectroscopy (FTIR) surface analysis. This work was performed by the IBM Research at various IBM Research and Development Facilities.

In agreement with the ITRS roadmap, there have been several publications supporting the reduction in critical dimensions and the introduction of new materials to semiconductor processing (1,2,3). This paper highlights the observations and solutions to some of the critical material and process interactions encountered during the integration of the back end of line interconnect.

We report FDSOI devices with a 20nm gate length (L G ) and 5nm spacer, featuring a 20% tensile strained Silicon-on-Insulator (sSOI) channel NFET and 35% [Ge] partially compressive strained SiGe-on-Insulator (SGOI) channel PFET. This work represents the first demonstration of strain reversal of sSOI by SiGe in short channel devices. At V dd of 0.75V, competitive effective current (I eff ) reaches 550/340 µA/µm for NFET, at I off of 100/1 nA/µm, respectively. With a fully strained 30% SGOI channel on thin BOX (20nm) substrate and V dd of 0.75V, PFET I eff reaches 495/260 µA/µm, at I off of 100/1 nA/µm, respectively. Competitive sub-threshold slope and DIBL are reported. With the demonstrated advanced strain techniques and short channel performance, FDSOI devices can be extended for both high performance and low power applications to the 10nm node.

Pattern collapse has long been known in photoresist patterning where the resist patterns merge or collapse during rinsing and drying steps [. The forces responsible for this collapse were identified as capillary forces during the drying process. Structures such as titanium nitride DRAM cylinders [ and silicon Flash shallow trench isolation (STI) lines have also been observed to be pattern collapse sensitive due to increase in aspect ratio of the features. Micro-electromechanical systems (MEMS) devices also show a similar phenomenon, but on a larger length scale, and is referred to as stiction [. For the technology nodes

A 10nm logic platform technology is presented for low power and high performance application with the tightest contacted poly pitch (CPP) of 64nm and metallization pitch of 48nm ever reported in the FinFET technology on both bulk and SOI substrate. A 0.053um 2 SRAM bit-cell is reported with a corresponding Static Noise Margin (SNM) of 140mV at 0.75V. Intensive multi-patterning technology and various self-aligned processes have been developed with 193i lithography to overcome optical patterning limit. Multi-workfunction (WF) gate stack has been enabled to provide Vt tunability without the variability degradation induced by channel dopants.

This work demonstrates the building of a 56 nm pitch copper dual damascene interconnects which connects to the local interconnect level. This M1/V0 dual-damascene used a triple pitch split bi-directional M1 and a double pitch split contact (V0) scheme where the local interconnects are with double pitch split in each direction, respectively. This scheme will provide great design flexibility for the advanced logic circuits. The patterning scheme is multiple negative tone development lithography-etch. A memorization layer is utilized in the triple patterned M1 and the double patterned V0 levels, respectively. After transferring the two via levels into the metal memorization layer, a self-aligned-via (SAV) RIE scheme was used to create vias confined by line trenches such that via to line spacing is maximized for better reliability. Seven litho/etch steps (LIP1/LIP2/V0C1/V0C2/M1P1/M1P2/M1P3) were employed to present this revolutionary interconnects.

This work demonstrates the building of 64 nm pitch copper single and dual damascene interconnects using pitch split double patterning scheme to enable sub 80nm pitch patterning. A self-aligned-via (SAV) litho/RIE scheme was used to create vias confined by line trenches such that via to line spacing is maximized for better reliability. An undercut free post RIE trench profile enabled the good metal fill. Initial reliability test result and the possibility of using the same scheme for 56 nm pitch interconnects are also discussed.

A self-aligned via(SAV) process was employed to build 64nm pitch Dual-Damascene(DD) interconnects using a pitch split double exposure pattering scheme to form the Cu lines. TiN hardmask (HM) density and thickness were optimized to achieve the SAV process and DD structure build. We present STEM cross sections of the structures after TiN HM deposition, HM open and DD RIE to determine the minimum required TiN HM thickness for the SAV process. We characterized the TiN loss for each RIE step from cross section results and defined the optimal TiN thickness for 64nm pitch interconnects. Using the optimized TiN thickness, we fabricated DD structures and compared the metal-to-via short electrical performance for SAV and non-SAV processes to show the overlay (OL) impact on shorts yield. Structures fabricated using the SAV process have excellent yield regardless of the degree of via misalignment in the SAV direction since no via CD growth occurs in the constrained SAV direction, while those processed with a non-SAV scheme show via yield degradation with increasing via misalignment. Also, with respect to misalignment in the non-SAV direction, there were no significant electrical differences between structures made using SAV and non-SAV approaches.

Underlayers (UL), such as organic planarizing layers (OPLs) or spin-on carbon (SOC) layers, play a very important role in various integration schemes of chip manufacturing. One function of OPLs is to fill in pre-existing patterns on the substrate, such as previously patterned vias, to enable lithographic patterning of the next level. More importantly, OPL resistance to reactive ion etch (RIE) processes used to etch silicon-containing materials is essential for the successful pattern transfer from the resist into the substrate. Typically, the pattern is first transferred into the OPL through a two-step RIE sequence, followed by the transfer into the substrate by a fluorine-containing RIE step that leaves the OPL pattern mainly intact. However, when the line/space patterns are scaled down to line widths below 35 nm, it was found that this last RIE step induces severe pattern deformation ("wiggling") of the OPL material, which ultimately prevents the successful pattern transfer into the substrate. In this work, we developed an efficient process to evaluate OPL materials with respect to their pattern transfer performance. This allowed us to systematically study material, substrate and etch process parameters and draw conclusions about how changes in these parameters may improve the overall pattern transfer margin.

Future microprocessor technologies will require interlayer dielectric (ILD) materials with a dielectric constant (κ-value) less than 2.5. Organosilicate glass (OSG) materials must be nanoporous to meet this demand. However, the introduction of nanopores creates many integration challenges. These challenges include 1) integrating nanoporous films with low mechanical strength into conventional process flows, 2) managing etch profiles, 3) processinduced damage to the nanoporous ILD, and 4) controlling the metal/nanoporous ILD interface. This paper reviews research to maximize mechanical strength by engineering optimal pore structures, controlling trench bottom roughness induced by etching and understanding its relationship to pore size, repairing plasma damage using silylation chemistry, and sealing a nanoporous surface for barrier metal (liner) deposition.

Low-pH titanium nitride (TiN) removal formulations utilizing oxidizers other than hydrogen peroxide (H2O2) have been developed with superior TiN selectivity toward W and the inter-level dielectric (ILD) films. One critical challenge is protecting the TiN barrier layer between the W and dielectric layer, which is often exposed during the TiN metal hard mask (MHM) removal step. This study focused on developing and optimizing formulations with TiN MHM etch rates ≥ 100 Å/min at temperatures ≤ 60oC, with compatibility toward W, ILD films, and the TiN liner. The galvanic corrosion was tailored to protect the TiN liner in the presence of W. A simple yet novel patterned wafer test vehicle was developed to facilitate this work, enabling the investigation of the chemical impact at the W/TiN liner interface.

not Available.

The resistance to plasma-induced damage of various nanoporous, ultra low-κ porous SiCOH films used as interconnect dielectric materials in integrated circuits was studied. These films are susceptible to damage by plasma processes used during nanofabrication. The dielectric constants and chemical compositions of four dielectric films were correlated with measured amounts of plasma damage. Films deposited with higher carbon content in the form of Si–CH3 and Si(CH3)2 bonding exhibited less plasma damage than similar films with lower carbon content.

The preparation of ultralow dielectric constant porous silicon, carbon, oxygen, hydrogen alloy dielectrics, called “pSiCOH,” using a production 200mm plasma enhanced chemical vapor deposition tool and a thermal treatment is reported here. The effect of deposition temperature on the pSiCOH film is examined using Fourier transform infrared (FTIR) spectroscopy, dielectric constant (k), and film shrinkage measurements. For all deposition temperatures, carbon in the final porous film is shown to be predominantly Si–CH3 species, and lower k is shown to correlate with increased concentration of Si–CH3. NMR and FTIR spectroscopies clearly detect the loss of a removable, unstable, hydrocarbon (CHx) phase during the thermal treatment. Also detected are increased cross-linking of the Si–O skeleton, and concentration changes for three distinct structures of carbon. In the as deposited films, deposition temperature also affects the hydrocarbon (CHx) content and the presence of CO and CC functional groups.