Tungsten Chemical Vapor Deposition Research Articles

Effect of Carrier Gases on the Chemical Vapor Deposition of Tungsten from WF 6 ‐ SiH4

Effect of carrier gases like argon, helium, and hydrogen on the deposition rate and properties of CVD tungsten from was studied. Each carrier gas has different momentum and thermal diffusivity, and the mean free path and mass diffusivity of reactant molecules are also different. This has a significant effect on the deposition rate, film composition, and film morphology. A carrier gas with high thermal diffusivity and high molecular velocity like helium lowers the deposition rate considerably because gas‐phase formations of intermediate species, which go out of the reactor, are favorable. The film looks like powdery aggregates and shows poor adhesion. Effect of the carrier gases on the step‐coverage, grain size, chemical composition, and electrical resistivity was identified from the experiment.

Cross-section TEM for examinations of selective tungsten CVD for via hole filling

The reaction between tungsten layers [deposited by chemical vapor deposition (CVD) as well as by physical vapor deposition (PVD)] and an aluminium alloy (AlSi) has been investigated. For CVD tungsten layers deposited on AlSi by silane reduction of WF6 the formation of an aluminium fluoride interlayer has been established by cross-section transmission electron microscopy (XTEM). At the interface between PVD-W and AlSi, an intermetallic compound may be formed depending on the thermal treatment. A crystalline intermediate layer has been found after annealing at 400°C (twice for 30 min), whereas an interface between as deposited double layers showed no interlayer. For all samples investigated, diffusion of aluminium into the tungsten layer was observed. The diffusion depth depends on the heat treatment and the tungsten morphology. Intermediate layers — if they exist — limit the diffusion but do not act as a diffusion barrier.

Selective Tungsten Chemical Vapor Deposition with High Deposition Rate for ULSI Application

For the application of selective tungsten chemical vapor deposition (CVD) on ULSI device fabrication, two serious problems still remain: loss of selectivity and damage on devices. Selective tungsten deposition with the high rate of 1 µm/min has been achieved by a new CVD system having a cold susceptor, where the high selectivity has been well maintained continuously. We have also found that the tungsten-Si contacts with good performance of low contact resistance and no device damage can be formed by controlling the initial growth of tungsten.

In Situ Monitoring of the Products from the SiH4 + WF 6 Tungsten Chemical Vapor Deposition Process by Micro‐Volume Mass Spectrometry

We have used a novel application of in situ high pressure mass spectrometry to monitor the tungsten chemical vapor deposition process. We observed the production of both and in ratios that vary with the inlet flow ratio. The and partial pressures were monitored over a tungsten coated wafer at 320°C as the inlet flow ratio was varied from a selective (0.25) to a nonselective (3.0) regime. We observed the product ratio to decrease dramatically at reactant ratios associated with the onset of loss of selectivity (1–1.5). The reaction products were sampled by means of a fused silica capillary that was introduced between the reacting tungsten coated wafer and a differentially pumped mass spectrometer. Gas phase reaction and thermodiffusion effects were also observed. These data provide insight concerning loss of selectivity mechanisms, and illustrate the potential of this mass spectrometric technique for real‐time in situ process control.

Kinetic model for the chemical vapor deposition of tungsten in the silane reduction process

The kinetics involved in the chemical vapor disposition of tungsten, using the silane reduction chemistry, is modeled by a triple Eley–Rideal mechanism, which includes three competitive surface reaction pathways: (1) silane reduction of surface adsorbed WF6, (2) surface dissociation of silane and silicon reduction of WF6, and (3) (residual) hydrogen reduction of surface adsorbed WF6. The last two reaction pathways have not been properly considered in the past, and will be shown to be of critical importance to the understanding the complex Arrhenius behavior exhibited by the silane reduction process. A quantitative deposition rate expression, with no fitting parameters, is derived from the kinetic model using kinetic constants in the literature. The model can predict the deposition rate obtained experimentally, after taking into account the effect of reactant depletion in the reactor.

Plasma enhanced chemical vapour deposition of tungsten and tungsten silicide thin films

The plasma enhanced chemical vapour deposition of tungsten and tungsten silicide thin films onto GaAs has been investigated over a range of deposition conditions using a modified parallel plate radial flow reactor. The tungsten films deposited from WF 6 and H 2 were typically smooth and adherent with as-deposited resistivities in the range 20–100 μΩ·cm and a grain size in the range 200–2000 Å. Schottky properties for diodes fabricated from the as-deposited layers showed current-voltage characteristics governed by thermionic emission with ideality factors ( n) as low as 1.04. WSi x films deposited from WF 6 and SiH 4 were typically smooth and adherent with a grain size <200 Å. Schottky diodes fabricated from these layers also exhibited good Schottky properties. It was possible to deposit films ranging from nearly pure tungsten to nearly pure silicon using this technique by varying the WF 6:SiH 4 gas flow ratio.

Reaction pathways for ArF excimer laser assisted tungsten chemical vapor deposition from a WF6–H2 gas mixture

Photolytic tungsten chemical vapor deposition by an ArF excimer laser from a WF6/H2/Ar gas mixture is a very complicated process with reactions occurring both in series and in parallel. In this article different reaction pathways were modeled and compared with experimental deposition rate data. The absorption cross section for WF6 was measured to 1.7×10−18 cm2 at the ArF excimer laser wavelength of 193 nm. From a rate equation model, it was concluded that the direct photolytic contribution to the deposition rate was negligible. Moreover, reduction of tungsten subfluorides by hydrogen atoms, formed in reactions between photolytically released fluorine atoms and molecular hydrogen, was also found to be negligible. However, photolysis in combination with various radical reactions generate a relatively high concentration of tungsten subfluorides (WF3, WF4, and WF5). Thermochemical calculations indicate that H2 reduction of tungsten subfluorides to solid tungsten, forming clusters in the vapor, is highly probable. Among the different cluster nucleation mechanisms polymerization of tungsten subfluorides seems to be an important step. This also explains the influence of WF6 partial pressure on the deposition rate.

The evolution of growth stresses in chemical vapor deposited tungsten films studied by in situ wafer curvature measurements

An in situ study of the evolution of the biaxial state of intrinsic stress during nucleation and growth of polycrystalline tungsten chemical vapor deposition films deposited by the hydrogen reduction of tungsten hexafluoride is presented. The evolution of biaxial stress was determined from in situ wafer curvature measurements. It is shown that the intrinsic stress is a growth stress, i.e., a stress developing in close vicinity to the advancing surface of the film due to metastable film growth processes. The stress developing depends strongly on the thickness of the film. High tensile stress (≊4 GPa) is observed during the initial stage of growth, compressive stress (up to −1 GPa) is observed in an intermediate thickness regime after film closure and tensile stress (0.1–1 GPa) is observed in the thick film regime. The associated stress gradients in the film are preserved during and after growth. The development of growth stress is determined by deposition temperature and growth rate. The tensile stress in the thick film regime is larger at a higher growth rate or a lower deposition temperature, while the compressive stress in the intermediate thickness regime showed the opposite dependency. Film properties as the evolution of grain size, impurity content, and resistivity are found not to vary significantly with the growth conditions. Therefore, the development of growth stress is ascribed to kinetical processes. The development of tensile stress in the thick film regime is described with a (kinetic) grain boundary formation and relaxation model. The compressive stress in the intermediate thickness regime is tentatively ascribed to compressive coherency strains induced by interfacial tensions of the grains in the stage of island growth.

Characterization of chemically vapor deposited tungsten on low-pressure chemically deposited and reactively sputtered titanium nitride films

The growth of low-pressure chemical vapor deposition (LPCVD) tungsten on LPCVD-grown and reactively sputtered titanium nitride (TiN) films has been studied employing the x-ray photoelectron spectroscopy depth profiling technique. Two types of gas ambients were used to grow tungsten (W) films at a deposition temperature of 400 °C. In one case, a very thin overlayer of W was formed on both types of TiN films by exposing the samples to WF6. The W thus deposited diffuses deep into the bulk of the reactively sputtered TiN film whereas it is confined to the W/TiN interface for the LPCVD-grown TiN film. In the second case, both types of TiN films were exposed to the WF6, SiH4, and H2 ambient. No diffusion of W into any of the TiN films was observed.

A Selective SiO2 Film‐Formation Technology Using Liquid‐Phase Deposition for Fully Planarized Multilevel Interconnections

A selective film‐formation technology using liquid‐phase deposition (LPD) around room temperature for fully planarized multilevel interconnections is developed. The LPD technique utilizes supersaturated hydrofluosilicic acid aqueous solution as a source liquid. The films can be selectively formed on chemical vapor deposition (CVD) underlayers in the trenches between photoresist patterns or tungsten wiring with photoresist as mask. For polysilicon patterns with photoresist masks, the films creep along the polysilicon and photoresist sidewalls. The selective deposition mechanism can be explained as siloxane oligomers, which are formed in the supersaturated aqueous solution, have different chemical reactivity between the photoresist and substrate surface. Global planarization of trenches between tungsten wiring is achieved, using the selective deposition technique. A fully planarized double‐level tungsten interconnection is realized using both selective film deposition and selective tungsten CVD via filling. Low contact resistance of ca. 0.3 Ω/unit is achieved for via holes 0.8 μm in diam.

X-ray photoelectron spectroscopic study of the chemical vapor deposited W/Al interface

The dependence of the amount of aluminum trifluoride (AlF3) piled up at the interface of chemical vapor deposited tungsten and the aluminum under layer on the deposition time and subsequent annealing in ultrahigh vacuum (UHV) or in monosilane (SiH4) gas has been studied. AlF3 is formed by the reaction of the aluminum under layer with tungsten hexafluoride (WF6) during the initial state of tungsten chemical vapor deposition. Tungsten was deposited on an Al layer under selective deposition conditions by SiH4 reduction at 250 °C. X-ray photoelectron spectroscopy measurement reveals that the amount of AlF3 decreases with an increase in the tungsten deposition time and that the reduction of AlF3 by volatilization of aluminum fluorides, which occurs at higher temperatures (≳400 °C) is not observed at low temperature (270 °C). Annealing in SiH4 gas after the tungsten deposition was effective to reduce the amount of AlF3 compared with annealing in UHV. This result and thermochemical data would suggest that the dependence of the amount of AlF3 on the tungsten deposition time is explained by the reduction of AlF3 with hydrogen atoms supplied from the dissociation of SiH4.

Silane reduced chemical vapor deposition tungsten as a nucleating step in blanket W

The SiH4–WF6 nucleation step of chemical vapor deposition blanket tungsten has been studied in the 10–100 Torr pressure range. The deposition rate, particle generation rate, and step coverage were studied as a function of the process parameters. The deposition rate was found to be mass transfer limited over the entire range of process conditions. Deposition rates from 300 to over 70 000 Å/min were observed. An in situ detector was used to measure gas phase particle levels. Particle generation rates were found to range from 0 to 108 particles/s over the parameter range studied. The strongest dependence seen was on the SiH4/WF6 flow ratio. At ratios of less than one, no detectable particles were produced. The particle generation rate increases rapidly at ratios above one. The particle rate also increases with increasing temperature and pressure. The step coverage was found to be ∼85% for 1000 Å films in 3:1 aspect ratio trenches. Overall, the results indicate that SiH4 nucleation in the 10–100 Torr range is an attractive alternative to the ≤1 Torr processes commonly in use today.

Chemical vapour deposition of tungsten in a UHV system: effect of substrate conditions on the initial stages of growth

The initial stages of growth in chemical vapour deposition of W from WF 6 have been studied in a UHV system using three different substrate conditions: HF-etched Si(100), partially oxidized Si(100) and an Si(100) substrate with a native oxide. A more detailed study of the initial growth process was possible using very low total pressures of WF 6(10 −6−10 −5 Torr). It was found that the interaction between WF 6 and the HF-etched substrate was affected strongly by a passivating layer of chemisorbed H-atoms on the Si surface. The hydrogenated surface reduced the nucleation rate and played an important role in keeping the WSi interface clean. The native oxide and the partially oxidized substrate were more reactive than the HF-etched substrate towards WF 6 over a broad temperature range. This effect was attributed to the presence of hydroxyl groups on the substrate surface.

Characterization of Encroachment of Selective Chemical Vapor Deposition of Tungsten at the Edge of Contacts

The encroachment of selective chemical vapor deposition of tungsten at the edge of contacts has been examined with RBS, SEM, stylus height measurements, and TEM. Encroachment is minimized by using reduction instead of reduction, a deposition temperature of 340°C or less, an in situ plasma clean, and a flow ratio ≥0.1. Use of a plasma clean can reduce the W grain size by a factor of four and reduce nonuniform consumption of the polysilicon. The improvements using a plasma clean are associated with an increase in the density of chemical vapor deposition of tungsten nucleation sites.

Modeling, optimization and control of spatial uniformity in manufacturing processes

Spatial uniformity, or the uniformity of product output characteristics at different locations in a batch of product is modeled, optimized, and controlled using a methodology called multiple response surfaces, which may be used to characterize the results of an experimental design. Multiple, low-order polynomial models are used to model the output characteristics at each of several sites within a batch of product. The uniformity model is then obtained by manipulating these multiple models. The approach is compared to the traditional method of fitting a single high-order polynomial to the calculated uniformity. Experimental results confirm that similar or improved modeling accuracy is obtained with fewer data points using the new method due to the use of low-order models. Characteristics of the approach are examined both analytically and in application to plasma etching, silicon epitaxy, tungsten chemical vapor deposition (CVD) and the simulation of polysilicon low-pressure CVD (LPCVD).< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

Metal CVD for microelectronic applications: An examination of surface chemistry and kinetics

Abstract We review the surface chemistry and kinetics relevant to the chemical vapor deposition (CVD) of metals used for microelectronic applications. Our efforts focus on the surface chemistry of aluminum, tungsten, and copper CVD, which have received the most recent interest for metallization. We first briefly review a variety of topics concerning the applications and the chemistry and kinetics of metal CVD. We also give a brief overview of the application of surface science techniques to the study of CVD-related surface chemistry.

Pulsed laser-assisted chemical vapor deposition of tungsten, molybdenum, and vanadium thin films

ADVERTISEMENT RETURN TO ISSUEPREVArticleNEXTPulsed laser-assisted chemical vapor deposition of tungsten, molybdenum, and vanadium thin filmsW. Turney, Y. M. Hung, S. G. Starcevich, P. S. Cardinahl, V. H. Grassian, and K. A. SingmasterCite this: Chem. Mater. 1992, 4, 6, 1192–1199Publication Date (Print):November 1, 1992Publication History Published online1 May 2002Published inissue 1 November 1992https://doi.org/10.1021/cm00024a017RIGHTS & PERMISSIONSArticle Views137Altmetric-Citations6LEARN ABOUT THESE METRICSArticle Views are the COUNTER-compliant sum of full text article downloads since November 2008 (both PDF and HTML) across all institutions and individuals. These metrics are regularly updated to reflect usage leading up to the last few days.Citations are the number of other articles citing this article, calculated by Crossref and updated daily. Find more information about Crossref citation counts.The Altmetric Attention Score is a quantitative measure of the attention that a research article has received online. Clicking on the donut icon will load a page at altmetric.com with additional details about the score and the social media presence for the given article. Find more information on the Altmetric Attention Score and how the score is calculated. Share Add toView InAdd Full Text with ReferenceAdd Description ExportRISCitationCitation and abstractCitation and referencesMore Options Share onFacebookTwitterWechatLinked InReddit PDF (2 MB) Get e-Alerts Get e-Alerts

Heat flow modeling of chemical vapor deposition of tungsten for improved film thickness uniformity

The boundary element method is used to model the heat flow in a tungsten chemical vapor deposition (CVD W) reactor to determine the origin of the tungsten film thickness variation across a 125 mmp-type silicon wafer. The typical thickness nonuniformity is 7-10%. The model predicts temperature variations in the graphite susceptor which heats the wafers. The resulting temperature difference between the center and edge of a wafer can be up to 20 °C. Deposition rate experiments in a CVD W reactor confirm this. Experiments involving changes in the gas flow geometry have been conducted. It is concluded that the temperature gradient across the wafer, not the gas flow geometry, is the dominant factor controlling film thickness uniformity for this CVD W reactor.

Organometallic chemical vapor deposition of tungsten metal, and suppression of carbon incorporation by codeposition of platinum

Highly reflecting, amorphous, thin films of tungsten are obtained by the decomposition of bis-cyclopentadienyltungstendihydride, (η-C5H5)2WH2, in 1 atm of hydrogen at 350 °C. Auger depth profiling reveals that the carbon and oxygen content of the films are 25.1 and 3.1 at. %, respectively. Simultaneous chemical vapor deposition of tungsten with a small amount of platinum reduces the carbon and oxygen content of the film to 5.3 and 1.8 at. %. The platinum is deposited from cyclopentadienylplatinumtrimethyl, (η-C5H5)Pt(CH3)3, and its concen- tration in the film is 3.3%. Annealing at 750 °C in hydrogen converts the tungsten into a polycrystalline deposit which exhibits an x-ray diffraction pattern characteristic of the metal. The sheet resistivities of the amorphous films are 52±4 μΩ cm.

Growth characteristics of photolytic laser-induced chemical vapor deposition of tungsten from WF6

ArF laser induced chemical vapor deposition (CVD) of W onto (100)Si from a WF6/H2/Ar mixture and with the laser beam in parallel to the substrate surface was studied. The influence of substrate temperature on the deposition rate was investigated in a broad range (25–450 °C) for both focused and unfocused beam conditions and compared with thermal CVD data. Above the threshold temperature for the thermal process, 250 °C, a combined thermal and photolytic process takes place. Below 250 °C the laser induced photolytic process yields deposition with a nearly temperature independent deposition rate of about 100 Å/min when a focused beam is used. For an unfocused beam a slightly lower deposition rate was observed. The distance between the laser beam and the substrate surface had only a slight influence on the deposition rate. For the focused beam conditions the deposition rate was only reduced by roughly 10% when the distance was increased from 0 to 8 mm. For the unfocused case the deposition rate increased, without reaching the focused beam values, upon increasing the beam to substrate distance in the range 0–8 mm. The substrate temperature had a strong influence on the resistivity of the deposited films for both laser-induced and thermal processes. For obtaining a low resistivity value, 8 μΩ cm, a substrate temperature above 350 °C was required. This was mainly due to the fact that the high-resistivity β phase was formed at lower temperatures. The influence of H2 concentration on the growth rate was investigated by replacing H2 with Ar. A small amount of H2 was enough to increase the deposition rate considerably. Use of Ar diluted gas resulted also in a higher deposition rate. Finally, the deposition rate increased linearly with increased total pressure.