Chemical vapor deposition of ruthenium films for metal electrode applications

Abstract

The present invention provides a method of depositing ruthenium films on a substrate via liquid source chemical vapor deposition wherein the source material is liquid at room temperature and utilizing process conditions such that deposition of the ruthenium films occurs at a temperature in the kinetic-limited temperature regime. Also provided is a method of depositing a thin ruthenium film on a substrate by liquid source chemical vapor deposition using bis-(ethylcyclopentadienyl) ruthenium by vaporizing the bis-(ethylcyclopentadienyl) ruthenium at a vaporization temperature of about 100-300° C. to form a CVD source material gas, providing an oxygen source reactant gas and forming a thin ruthenium film on a substrate in a reaction chamber using the CVD source material gas and the oxygen source reactant gas at a substrate temperature of about 100-500° C.

Description

BACKGROUND OF THE INVENTION

[0001] 1. Field of the Invention

[0002] The present invention generally relates to the liquid source chemical vapor deposition of metal films as the electrode on capacitors in integrated circuits. More particularly, bis(ethylcyclopentadienyl) ruthenium is used as the liquid source precursor for chemical vapor deposition of ruthenium films at a low temperature in the kinetic-limited temperature regime as the electrode for the application of metal-insulator-metal capacitors.

[0003] 2. Brief Description of the Related Art

[0004] In metal-insulator-metal capacitors, ruthenium is a preferred electrode material for next-generation dynamic random access memories (DRAMs). As the next generation DRAM technology evolves, it will become necessary to use three-dimensional capacitors despite the high-dielectric constant materials employed. The most important characteristics of a capacitor are high dielectric constant and small leakage current even when the electrode film is thin. If the films are being formed for electrodes for DRAM applications, it is advantageous to use a chemical vapor deposition process (CVD).

[0005] Chemical vapor deposition is a broad class of processes using controlled chemical reactions to create layers on wafers and is a key process in the development of ultra-large-scale integrated circuit fabrication. Chemical vapor deposition of thin metal films realizes good step coverage and wafer-to-wafer repeatability on complicated topography. However, the source material for CVD processes must be stable and possess good vaporization properties.

[0006] The use of liquid source metalorganic precursors for the CVD formation of thin films provides a means of repeatably creating these wafers. CVD precursors for advanced DRAM electrodes and dielectrics have traditionally been solid compounds and although soluble in organic solvents such as tetrahydrofuran, solubility is limited. CVD deposition of films can require high vaporization temperatures and residues are left after vaporization and deposition. Barium strontium titanate thin metal films on a substrate with a BST sputtered film using titanyl bis (dipivaloymethanto) (TIO(DPM)) and titantium tetraisopropoxide (TTIP) organometallic solutions as the CVD liquid source material have been formed.

[0007] Traditionally, bis(cyclopentadienyl) ruthenium (Ru(Cp)2) is used to deposit thin ruthenium films on substrates. However, this precursor is a solid at room temperature possessing the concomitant problems of limited solubility in tetrahydrofuran, lower deposition rate and remaining residue after vaporization which could cause particle formation, process drift and prevent precursor transport. Additionally, bis(cyclopentadienyl) ruthenium easily forms a ruthenium oxide film on substrates at low temperatures in the kinetic-limited temperature regime for virtually all process conditions.

[0008] It is therefore, advantageous, to use a liquid source for CVD thin film deposition that is a stable liquid at room temperature, yet is less susceptible to oxidation during vaporization and deposition of thin films. Ruthenium films have been deposited on silicon substrates using bis-(ethylcyclopentadienyl) ruthenium. However, these ruthenium films had a large column width and contained large quantities of carbon and hydrogen impurities which resulted in a resistivity greater than that of films deposited using the solid precursor bis-(cyclopentadienyl) ruthenium, although, the resistivity was still sufficiently small to be used as capacitor electrodes.

[0009] Bis-(alkyl-cyclopentadienyl) ruthenium complexes, including Ru(EtCp)2, have been synthesized and have been used in a process to produce ruthenium-containing films. These pure ruthenium films were deposited on a silicon substrate at 600° C. in a hydrogen atmosphere. Thus, deposition occurred in the mass-transfer limited regime.

[0010] The prior art is deficient in the lack of effective means of depositing pure thin ruthenium films on a substrate using a chemical vapor deposition source that is liquid at room temperature and where deposition temperature occurs in the kinetic-limited temperature range. The present invention fulfills this long-standing need and desire in the art.

SUMMARY OF THE INVENTION

[0011] The present invention provides a method of depositing ruthenium films on a substrate via liquid source chemical vapor deposition wherein the source material is liquid at room temperature and utilizes process conditions such that deposition of the ruthenium films occurs at a temperature in the kinetic-limited temperature regime.

[0012] In another embodiment of this invention, there is provided a method of depositing a thin ruthenium film on a substrate by liquid source chemical vapor deposition, using bis-(ethylcyclopentadienyl) ruthenium, by vaporizing the bis-(ethylcyclopentadienyl) ruthenium at a vaporization temperature of about 100-300° C. to form a CVD source material gas and forming a thin ruthenium film on a substrate in a reaction chamber using the CVD source material gas and the oxygen source reactant gas such that the substrate has a temperature of about 100-500° C. and deposition of the ruthenium film occurs in the kinetically limited temperature regime.

[0013] Other and further aspects, features, and advantages of the present invention will be apparent from the following description of the presently preferred embodiments of the invention given for the purpose of disclosure.

BRIEF DESCRIPTION OF THE DRAWINGS

[0014] So that the matter in which the above-recited features, advantages and objects of the invention, as well as others which will become clear, are attained and can be understood in detail, more particular descriptions of the invention briefly summarized above may be had by reference to certain embodiments thereof which are illustrated in the appended drawings. These drawings form a part of the specification. It is to be noted, however, that the appended drawings illustrate preferred embodiments of the invention and therefore are not to be considered limiting in their scope.

[0015] FIG. 1 depicts the Thermal Gravimetric Analysis and Differential Scanning Calorimetry data for the Ru(Cp)2 (FIG. 1A) precursor the Ru(EtCp)2 (FIG. 1B) precursors.

[0016] FIG. 2 depicts a schematic of an MLDS dual-vaporizer Ru chamber used for the chemical vapor deposition of ruthenium films. Chemical vapor deposition ruthenium deposition only uses one vaporizer.

[0017] FIG. 3 depicts the Vaporizer Temperature Optimization Curve for the Ru(Cp)2 (FIG. 3A) precursor the Ru(EtCp)2 in octane (FIG. 3B) precursor.

[0018] FIG. 4 depicts the Arrhenius plot of CVD Ru process using Ru(Cp)2 precursor on 200 Å CVD TiN and on 2000 Å thermal oxide (FIG. 4A) and 60 Å PVD Ru seed layer (FIG. 4B) and using Ru(EtCp)2 in octane on 200 Å PVD TiN (FIG. 4C) and 60 Å PVD Ru seed layer (FIG. 4D).

[0019] FIG. 5 depicts the Arrhenius plot of CVD Ru process using pure Ru(EtCp)2 at different Ru and O2 flow rates with constant N2 flow rate of 1300 sccm on a 50 Å PVD Ru seed layer/2 kÅ ThOx substrate. The activation energies are also shown. N2 carrier flow is 450 sccm.

[0020] FIG. 6 depicts the Arrhenius plot of CVD Ru process deposition rate and resistance uniformity using pure Ru(EtCp)2 at flow rates of 50 mgm Ru and 100 sccm O2 (FIG. 6A) and rates of 20 mgm Ru and 40 sccm O2 (FIG. 6B) on a 50 Å PVD Ru seed layer/2 kÅ ThOx substrate. N2 flow is 1300 sccm and N2 carrier flow is 450 sccm.

[0021] FIG. 7 depicts the Arrhenius plot of CVD Ru process deposition rate using pure Ru(EtCp)2 at N2 flow rates of 250 sccm, and 1300 sccm on a 50 Å PVD Ru seed layer/2 kÅ ThOx substrate. Ru flow is 50 mgm and N2 carrier flow is 450 sccm.

[0022] FIG. 8 depicts the Arrhenius plot of CVD Ru process deposition rate and resistivity using pure Ru(EtCp)2 at a high N2 flow rate of 1300 sccm on a 50 Å PVD Ru seed layer/2 kÅ ThOx substrate. Ru flow is 50 mgm, O2 flow rate is 100 sccm and N2 carrier flow is 450 sccm.

[0023] FIG. 9 depicts the X-Ray Diffraction (XRD) pattern of chemical vapor deposition Ru film grown with Ru(EtCp)2 in octane with a deposition temperature of 330° C., substrate of 60 Å physical vapor deposition Ru/2000 Å SiO2 and Ru thickness of 600 Å (FIG. 9A) and with a deposition temperature of 320° C., substrate of 60 Å PVD Ru/2000 Å SiO2 and Ru thickness of 280 Å (FIG. 9B) and with Ru(Cp)2 in tetrahydrofuran with a deposition temperature of 340° C., substrate of 60 Å PVD Ru/200 Å PVD TiN and Ru thickness of 800 Å (FIG. 9C), with a deposition temperature of 320° C., substrate of 60 Å PVD Ru/200 Å PVD TiN and Ru thickness of 650 Å (FIG. 9D) and, in the kinetic-limited regime, with a deposition temperature of 295° C., substrate of 60 Å PVD Ru/200 Å PVD TiN and Ru thickness of 500 Å (FIG. 9E).

[0024] FIG. 10 depicts the X-Ray Diffraction (XRD) pattern of chemical vapor deposition Ru film grown with Ru(EtCp)2 in octane on a PVD Ru seed substrate on 2000 Å thermal oxide and PVD Ru seed substrate on 200 Å PVD TiN.

[0025] FIG. 11 depicts the effect of O2 flow on CVD Ru orientation, using pure Ru(EtCp)2 on a 50 Å PVD Ru seed layer/2 kÅ ThOx substrate. Temperature is 355° C., N2 flow rate is 250 sccm and N2 carrier flow rate is 450 sccm.

[0026] FIG. 12 depicts the effect of O2 flow on CVD Ru (002) orientation and resistivity using pure Ru(EtCp)2 on a 50 Å PVD Ru seed layer/2 kÅ ThOx substrate. Process conditions are: 355° C., N2 250 sccm, N2 carrier 450 sccrn, and 8 torr.

[0027] FIG. 13 depicts SEM Images of the CVD Ru films to compare the surface roughness Ru(EtCp)2 precursor at 343° C. on 200 Å PVD TiN (FIG. 13A), Ru(EtCp)2 precursor at 343° C. on PVD Ru seed/2000 Å Oxide (FIG. 13B), Ru(Cp)2 precursor at 343° C. on PVD Ru seed/200 Å PVD TiN (FIG. 13C), Ru(EtCp)2 precursor at 343° C. on PVD Ru seed/2000 Å Oxide (FIG. 13D), Ru(Cp)2 precursor at 320° C. on PVD Ru seed/200 Å PVD TiN (FIG. 13E), Ru(EtCp)2 precursor at 330° C. on PVD Ru seed/2000 Å Oxide (FIG. 13F).

[0028] FIG. 14 depicts an AFM 2 &mgr;m×2 &mgr;m scan of CVD Ru film deposited using Ru(Cp)2 at 343° C. on PVD Ru seed/200 Å PVD TiN and PVD TiN.

[0029] FIG. 15 depicts the comparison of step coverage using Ru(Cp)2 precursor at two different temperatures on 200 Å PVD TiN layer. FIG. 15A: Bottom/Top Ratio ˜30%, Sidewall/Top Ratio ˜40%, Conformality ˜60%. FIG. 15B: Bottom/Top Ratio ˜90%, Sidewall/Top Ratio ˜90%, Conformality ˜80%. FIG. 15C reveals a roughness problem at larger geometry.

[0030] FIG. 16 depicts the comparison of step coverage using Ru(Cp)2 (FIG. 16A) and Ru(EtCp)2 in octane (FIG. 16B) on 200 Å PVD TiN seed layers at 340° C. with 0.17 &mgr;m AR 5:1. FIG. 16A: Bottom/Top Ratio ˜35%, Sidewall/Top Ratio ˜30%, Conformality ˜60%.

[0031] FIG. 17 depicts the comparison of step coverage using Ru(EtCp)2 in octane precursor with and without 60 Å Ru seed layer. FIG. 17A: Bottom/Top Ratio ˜30%, Sidewall/Top Ratio ˜30%, Conformality ˜50%. FIG. 17B: Bottom/Top Ratio 90%, Sidewall/Top Ratio ˜90%, Conformality ˜80%. Deposition temperatures were changed slightly due to CVD Ru process substrate sensitivity.

[0032] FIG. 18 depicts the comparison of step coverage using Ru(Cp)2 and Ru(EtCp)2 in octane on 60 Å PVD Ru seed layer. FIG. 18A: Bottom/Top Ratio ˜40%, Sidewall/Top Ratio ˜50%, Conformality ˜80%. FIG. 18B: Bottom/Top Ratio 90%, Sidewall/Top Ratio ˜90%, Conformality ˜80%.

[0033] FIG. 19 depicts the step coverage of a chemical vapor deposition Ru film using pure Ru(EtCp)2 at 355° C. for a 0.15 um trench, AR 6:1. Topfilm thickness 400 Å, side film thickness ˜350 Å, bottom film thickness ˜350 Å, sidewall coverage ˜88%, bottom coverage ˜88%.

[0034] FIG. 20 depicts a demonstration of top electrode application: CVD Ru deposited at 340° C. with Ru(EtCp)2 in octane precursor on 80 Å Ta2O5/1500 Å poly-Si substrate. CVD Ru film thickness ˜1500 Å. FIG. 20A: 0.1 um AR 16:1, Bottom/Top Ratio ˜10%, Sidewall/Top Ratio ˜10%, Conformality ˜40%. FIG. 20B: 0.05 um AR 20:1, Bottom/Top Ratio N/A, Sidewall/Top Ratio ˜5%, Conformality ˜25%.

[0035] FIG. 21 depicts SEM Images of CVD Ru films deposited on CVD BST/PVD Pt on, patterned cup structures.

[0036] FIG. 22 depicts process repeatability for Ru(Cp)2 and Ru(EtCp)2 in octane precursors. FIG. 22A: Repeatability plot for 100-wafer run using Ru(Cp)2 precursor. The wafer temperature was 340° C., and the substrate was 200 Å PVD TiN/Si. FIG. 22B: Repeatability plot for 25-wafer run using Ru(EtCp)2 precursor. The wafer temperature was 330° C., and the substrate was 60 Å PVD Ru/2000 Å thermal SiO2.

[0037] FIG. 23 depicts the effects of using a clamp ring for edge exclusion during CVD Ru deposition. FIG. 23A shows the positioning of the clamp ring; FIG. 23B is a thickness profile by alpha step of the CVD Ru film; FIG. 23C shows the SEM of the cross section of the wafer edge where the clamp ring is positioned; and FIG. 23D shows the presence of particles at the wafer's edge under the clamp ring.

[0038] FIG. 24 depicts the dependence of sheet resistance on film thickness.

DETAILED DESCRIPTION OF THE INVENTION

[0039] In one embodiment, the present invention provides a method of depositing ruthenium films on a substrate via liquid source chemical vapor deposition wherein the source material is liquid at room temperature and utilizes process conditions such that deposition of the ruthenium films occurs at a temperature in the kinetic-limited temperature regime. A representative example of the liquid source is bis-(ethylcyclopentadienyl) ruthenium. Representative examples of the substiate are thermal oxide, titanium nitride, titanium aluminum nitride, tantalum pentoxide, tantalum nitride, tantalum, barium strontium titanate, strontium oxide, ruthenium oxide, silicon nitride, tungsten nitride, lead zirconium titanate (PZT), strontium bismuth tantalate (SBT), and silicon dioxide. Additionally, the substrate wafers have a first film or seed layer deposited via vapor deposition, for example, physical vapor deposition or chemical vapor deposition. Representative examples of the seed layer are ruthenium, iridium, platinum, titanium nitride, titanium aluminum nitride, tantalum pentoxide, ruthenium oxide and titanium silicide. Also provided are process conditions comprising a ruthenium vaporization temperature of about 100-300° C. and a substrate temperature of about 100-500° C.

[0040] Another embodiment of the present invention provides a method of depositing a thin ruthenium film on a substrate by liquid source chemical vapor deposition, using bis-(ethylcyclopentadienyl) ruthenium as the liquid source. This method comprises vaporizing the bis-(ethylcyclopentadienyl) ruthenium at a vaporization temperature of about 100-300° C. to form a chemical vapor deposition source material gas, providing an oxygen source reactant gas and forming a thin ruthenium film on a substrate in a reaction chamber using the CVD source material gas and the oxygen source reactant gas such that the substrate has a temperature of about 100-500° C. and deposition of the ruthenium film occurs in the kinetically limited temperature regime. Representative examples of the substrate are thermal oxide, titanium nitride, titanium aluminum nitride, tantalum pentoxide, tantalum nitride, tantalum, barium strontium titanate, strontium oxide, ruthenium oxide, silicon nitride, tungsten nitride, lead zirconium titanate, strontium bismuth tantalate, and silicon dioxide. Additionally, the substrate wafers have a first film or seed layer deposited via vapor deposition, for example, physical vapor deposition or chemical vapor deposition. Representative examples of the seed layer are ruthenium, iridium, platinum, titanium nitride, titanium aluminum nitride, tantalum pentoxide, ruthenium oxide and titanium silicide.

[0041] The following definitions are given for the purpose of understanding the present invention. Any terms not expressly defined herein should be given their clear and ordinary meaning in the art.

[0042] As used herein the term, “seed layer” shall refer to a layer of material used to facilitate the growth of another layer in order to have desired properties; e.g., high nucleation density, low roughness and specific orientation.

[0043] As used herein the term, “kinetically-limited temperature regime” shall refer to the range of deposition temperatures over which the deposition rate of a chemical vapor deposition film is limited by the kinetics of the chemical reactions at the substrate surface, typically characterized by a strong dependence of deposition rate on temperature.

[0044] As used herein the term, “mass-transfer limited regime” shall refer to the range of deposition temperatures over which the deposition rate of a CVD film is limited by the flux of chemical reactants to the substrate surface, characterized by a strong dependence of deposition rate on chemical flow rates and independent of deposition temperature.

[0045] Process conditions for CVD deposition of ruthenium films are selected so that the ruthenium films are formed at lower temperatures in the kinetically limited temperature regime thus yielding pure ruthenium films possessing low oxygen content and low resistivity. Optionally, the substrate has a seed layer which controls the CVD ruthenium film roughness, improves the adhesion to the substrate and improves the step coverage and conformality on patterned structures.

[0046] In determining the optimum deposition conditions, the CVD ruthenium films are deposited using a range of process conditions for all precursors. The following film properties are compared: crystalline orientation, thickness WIW uniformity, resistivity, Rs WIW uniformity, surface roughness, step coverage, and conformality on patterned structures. X-ray diffraction (XRD) analysis of crystalline orientation shows that films deposited with the Ru(Cp)2 precursor in the kinetic-limited regime are significantly oxidized to RuO2, while films deposited with Ru(EtCp)2, either pure or 1.0 M in octane, in the same regime still contain significant amounts of pure ruthenium. For thickness within-wafer uniformity, there are metrology issues with the XRF and Metapulse tools, Ru(EtCp)2 in octane and Ru(Cp)2 are comparable with regard to resistivity and Rs WIW uniformity. SEM images in conjunction with AFM data show that CVD Ru films have high roughness whether deposited with Ru(EtCp)2 or Ru(Cp)2.

[0047] The use of a PVD Ru seed layer, approximately 50-60 Å thick, reduces surface roughness. Step coverage results show that conformal CVD Ru films in aggressive structures can be grown using Ru(EtCp)2 in octane or pure Ru(EtCp)2 or Ru(Cp) precursors. Both Ru(Cp)2 and Ru(EtCp)2 precursors demonstrate 90% step coverage (sidewall/top and bottom/top) and 80% conformality (sidewall thickness variation) for a 0.17 &mgr;m trench with a 5:1 aspect ratio (AR). Pure Ru(EtCp)2 precursor demonstrates ˜88% sidewall coverage and ˜88% bottom coverage for a 0.15 &mgr;m trench with a 6:1 (AR). From the demonstrable chemical properties, process performance, and film properties, Ru(EtCp)2 is an excellent precursor for CVD Ru process o n the basis of high deposition rate, low residue, good step coverage and wafer-to-wafer repeatability, and oxidation resistance in the kinetic-limited temperature regime.

[0048] The following examples are given for the purpose of illustrating various embodiments of the invention and are not meant to limit the present invention in any fashion.

EXAMPLE 1

[0049] Materials

[0050] Both bis(cyclopentadienyl) ruthenium (Ru(Cp)2) and bis(ethylcyclopentadienyl) ruthenium (Ru(EtCp)2) were manufactured and supplied by Advanced Chemical Delivery Systems (ACDS). These precursors are used to deposit CVD Ru films as the electrode for the application of metal-insulator-metal (MIM) capacitors. In general, the CVD Ru film must have the following characteristics: thickness non-uniformity within wafer less than 2% (1&sgr;, 49 pts/9 pts, 15 mmEE); wafer-to-wafer repeatability less than 2% (1&sgr;); film resistivity &rgr; less than 25&mgr;&OHgr;-cm; and sheet resistance non-uniformity less than 10% (1 s, 49 pts, 5 mmEE); and step coverage and sidewall conformality greater than 90%. The film needs to be reasonably smooth so that the capacitor stack is reliable and the film is continuous at thicknesses of 300 Å and below. Table 1 lists the target specifications together with the current process performance and best results for CVD Ru using Ru(EtCp)2. 1 TABLE 1 CVD Ru Current Results Current Best Target Known Method Thickness 300-1000 Å 300 Å Deposition Rate >100 Å/min >100 Å/min Thickness Uniformity WIW (9 pts. XRF)  <2%, 1&sgr; 3-4%, 1&sgr; WTW-  <2%, 1&sgr;  ˜2%, 1&sgr;(5 wafers) Resistivity 25 &mgr;&OHgr;-cm 22 &mgr;&OHgr;-cm Rs Uniformity 48 pts, 3 mm e.e. <10%, &sgr;  5%, 1&sgr; Conformality ≧90% ≧90% 0.15 &mgr;m @ 4:1, 300 Å Step Coverage N/A ˜90% Roughness Measure, TBD Haze 30-50 ppm no PVD Ru seed ->requires PVD seed Adhesion no delamination no delamination Particles <0.2P/cm2, <0.16 &mgr;m <0.2P/cm2, >0.16 &mgr;m*** C Content TBD MIM Capcitor Toxeq  <10 Å TBD J 1E-8A/cm2 @ 1V TBD * Wafer Y0f02603 run at 335° C. wafer temp. not BKM 320° C., no edge exclusion ** 25-wafer at Ru 180 mgm/O2 300 sccm/8/ Toor/N2 250 sccm 320° C. wafer/200 Å thickness ***Limited particle test on Si wafer after 650 wafers run through chamber

EXAMPLE 2

[0051] Physical Properties of Ru(Cp)2 and RU(EtCp)2

[0052] The Ru(CP)2 precursor is a solid at room temperature and is dissolved in the solvent Tetrahydrofuran (THF) in which the maximum allowable solubility is only 0.12 M. The Ru(EtCp)2 precursor is a liquid at room temperature and is dissolved in octane to 1.0M concentration or used in pure form. Table 2 compares the basic properties of the two precursors. 2 TABLE 2 Basic Properties of Ru(Cp)2 and Ru(EtCp)2 AMAT TGA Vapor- Current State @ Residue izer Precursor Precursor 25° C. M.P (ADCS) Temp. Solution Comments Ru(Cp)2 Solid 195° C. 2.5% (1 220° C. 0.1 M in Low atm O2) THF solubility limits dep. Rate Ru(EtCp)2 Liquid  12° C. No 260° C. 1.0 M in residue octane or 1 atm pure Ar

EXAMPLE 3

[0053] Thermal Gravimetric Analysis (TGA) and Differential Scanning Calorimetry (DSC)

[0054] A combination of Thermal Gravimetric Analysis (TGA) and Differential Scanning Calorimetry (DSC) was used to evaluate the precursors Ru(Cp)2 and Ru(EtCp)2 (FIGS. 1A and 1B). These measurements are performed at atmospheric pressure. The thermal gravimetric analysis curve measures the sample mass as a function of temperature. A decrease in sample mass occurs during evaporation, sublimation, or decomposition. The mass of the involatile residue remaining after thermolysis indicates the amount of residue that is left after vaporization. When the precursor has no residue, the process has less potential for particles and process drifts. The Ru(Cp)2 precursor has a residue of <3 wt % while Ru(EtCp)2 has close to 0 wt % residue.

[0055] Differential Scanning Calorimetry (DSC) curve measures the relative amount of heat flow from the sample with respect to a reference material. The differential scanning calorimetry curve shows the temperatures at which endothermic and exothermic transitions take place. The endothermic behavior indicates sublimation, evaporation or melting while exothermic reactions indicate reactions or decomposition. Thus, from the differential scanning calorimetry curve, the temperature range at which decomposition and sublimation occur are estimated. The information from the curve provides a starting point at which to set the vaporizer temperature. Ru(Cp)2 indicates an endothermic reaction (melting point) at 200° C. while a very broad endothermic peak is observed near 240° C. for Ru(EtCp)2.

[0056] Ru(EtCp)2 does not have limited solubility in solvents and can be used as a neat (pure) liquid or in a highly concentrated solution with a simple organic solvent, e.g., octane or tetrahydrofuran. As a result, the deposition rate is not limited as it is for the Ru(Cp)2 precursor. Thermal gravimetric analysis data shows that Ru(EtCp)2 leaves virtually no residue in the vaporizer or heated lines which could potentially lead to particles and process drift. Also, because Ru(EtCp)2 is a liquid at room temperature, it is less likely to leave solid residues which can block precursor transport through the liquid lines.

EXAMPLE4

[0057] Hardware and Process Conditions

[0058] All data was collected on an Applied Materials Ru chamber with the following configuration (FIG. 2):

[0059] Chamber: Ru

[0060] Faceplate: Standard

[0061] Heater: NGK HA-12 (95%)

[0062] Vaporizer: MLDS dual vaporizer (No precursor flow through 2nd vaporizer)

[0063] Frit: Dual, 100 &mgr;m pore size

[0064] Tip: 10 mil tapered to 4 mil, 0.1″ insertion length into top frit

[0065] Ampule Push Gas: 65 psi

[0066] The process conditions, unless specified otherwise, for Ru(Cp)2 and Ru(EtCp)2 comparisons were:

[0067] Ru(Cp)2 (0.1 M in tetrahydrofuran)

[0068] Heater Temp. 330-350° C.; Vaporizer Temp. 220° C. Lid/Liner/Nose Temp. 190° C.; Ru Flow 300 mgm; O2 flow 300 sccm; N2—B Ru 450 sccm; N2-B Pt 250 sccm. (2nd vaporizer); ˜Pressure 8 Torr; and D/R ˜60 Å/min at 340° C. on PVD Ru seed

[0069] Ru(EtCp)2 (1 M in octane):

[0070] Heater Temp. 330-350° C.; Vaporizer Temp. 260° C.; Lid/Liner/Nose Temp. 260° C.; Ru flow 180 mgm; O2 flow 300 sccm; N2—B 450 sccm (1st vaporizer); N2—B 250 sccm (2nd vaporizer); Pressure 8 Torr; and D/R ˜100 Å/min at 330° C. on PVD Ru seed

[0071] Ru(EtCp)2(pure):

[0072] Heater Temp. 330-350° C.; Vaporizer Temp. 260° C.; Lid/Liner/Nose Temp. 220° C.; Ru flow 50 mgm; O2 flow 100 sccm; N2 Flow 1300 sccm (to chamber); N2—B 450 sccm (1 vaporizer only); Pressure 2 Torr; and D/R ˜450 Å/min at 330° C. o n PVD Ru seed

EXAMPLE 5

[0073] Vaporizor Temperature Optimization

[0074] In order to evaluate the performance of Ru(Cp)2 and Ru(EtCp)2 on the Applied Materials Ru Captiva tool, vaporizer temperature optimization was performed. The Ru deposition rate was plotted against the vaporizer temperature (FIGS. 3A and 3B).

[0075] Based on these temperature optimization curves, a temperature of 220° C. for the vaporizer using Ru(Cp)2 and 260° C. using Ru(EtCp)2 was selected. These respective temperatures yield near the maximum deposition rate and good wafer-to-wafer repeatability for the respective precursors. The heated lines from the vaporizer to the chamber have the same temperatures. For Ru(Cp)2, however, the lid, liner, and nose temperatures were lowered, to 190° C. as a precautionary measure due to observed reaction with O2 in the showerhead area at a temperature of 250° C.

EXAMPLE 6

[0076] Kinetic Study Using Ru(Cp)2 (0.1 M in tetrahydrofuran) and Ru(EtCp)2 (1.0 M in octane)

[0077] The deposition temperature was varied for both Ru(Cp)2 and Ru(EtCp)2 precursors to determine the temperature dependence of the deposition rate using Arrhenius plots (log deposition rate vs. 1/T). Arrhenius plots are used to understand the reaction kinetics and to determine the temperature sensitivity of the CVD process. For good step coverage and conformality, the CVD process needs to be operated below the “knee” in the kinetic-limited regime. The “knee” is defined as the transition point between the mass-transfer limited regime and the kinetic-limited regime. Higher activation energies are generally not desirable for a stable metal organic (MOCVD) process due to high temperature sensitivity. The temperature dependence of the CVD Ru process for both Ru(Cp)2 and Ru(EtCp)2 precursors varies on different substrates thereby entailing an Arrhenius study on every substrate used.

[0078] The four substrates used for these Arrhenius studies are 2000 Å thermal oxide, 200 Å CVD TiN, 200 Å PVD TiN, and 60 Å PVD Ru seed layer. The CVD Ru process using either Ru(Cp)2 or Ru(EtCp)2 precursor is sensitive to substrate, so deposition rate and film properties vary on different substrates. Initially, thermal oxide substrates were used as substrates, but adhesion problems (peeling) and poor film quality and appearance were observed.. For the CVD Ru(Cp)2 process, CVD TiN and PVD TiN substrates improved the film adhesion and appearance. However, for CVD Ru(EtCp)2, the haze and surface roughness were found to be very high on the TiN substrates. The use of a PVD Ru seed layer, approximately 60Å thick, dramatically improved the haze of the CVD Ru(EtCp)2 films.

[0079] FIG. 4A shows the temperature dependence of the deposition rate for Ru(Cp)2 deposited on PVD TiN and thermal oxide substrates. The CVD Ru process for this Arrhenius plot uses slightly different conditions (e.g. 2 Torr pressure, O2 flow 500 sccm). The deposition rate dependence was similar on thermal oxide and CVD TiN. The activation energy is about 2-4 eV for Ru(Cp)2 on TiN and thermal oxide substrates, and the “knee” was close to 300° C.

[0080] The Arrhenius plot of Ru(Cp)2 on a PVD Ru seed layer is shown in FIG. 4B. The estimated activation energy of ˜0.7 eV was approximately four times lower than the activation energy of Ru(Cp)2 on TiN. A lower activation energy improves the process stability in the kinetic-limited regime. The temperature dependence of film resistivity is also plotted in FIG. 4B. The resistivity increases from about 20&mgr;&OHgr;-cm to 60&mgr;&OHgr;-cm when the temperature is reduced from 330° C. to 308° C. The sudden increase in resistivity is likely due to the formation of RuO2 at the lower temperatures as indicated in XRD analysis. There appears to be a transition point around 320° C. in which RuO2 is preferentially formed.

[0081] The Arrhenius plot of Ru(EtCp)2 on PVD TiN shown in FIG. 4C has an activation energy of ˜5 eV which is higher than Ru(Cp)2. Also, the curve and “knee” are shifted to slightly higher temperatures than the Ru(Cp)2 precursor. The higher activation energy is perhaps caused by the higher deposition rates achieved using the Ru(EtCp)2 precursor.

[0082] The Arrhenius plot and resistivity dependence for the Ru(EtCp)2 precursor on PVD Ru seed substrates is shown in FIG. 4D. The curve was shifted to higher temperatures relative to the Arrhenius plot for Ru(Cp)2 on PVD Ru seed substrates. The estimated activation energy for Ru(EtCp)2 over the observed temperature range was about 3 eV, approximately four times higher than the activation energy for Ru(Cp)2 on PVD Ru seed.

[0083] The resistivity increases gradually to 40&mgr;&OHgr;-cm as the deposition temperature was lowered, in contrast to the abrupt increase in resistivity observed with the Ru(Cp)2 precursor. In this case, the decrease in resisitvity may be caused by several factors: smaller grain size, decrease in film thickness, and some initial RuO2 formation. However, Ru(EtCp)2 appears to form less RuO2 than Ru(Cp)2 in the kinetic-limited regime.

[0084] The study of Arrhenius behavior of CVD Ru(Cp)2 and CVD Ru(EtCp)2 shows both processes to be sensitive to substrate. For both Ru(Cp)2 and Ru(EtCp)2 precursors, the activation energy was lower when deposited on a PVD Ru seed layer rather than on other substrates such as TiN and oxide. In addition, the deposition rates on PVD Ru seed layers are higher than on TiN, BST, or TanOx (latter two results not shown) in the kinetic-limited regime. The lower activation energies and higher deposition rates on PVD Ru highlight the importance of using a PVD Ru seed layer as a substrate for CVD Ru growth.

[0085] The Ru(EtCp)2 process has a higher activation energy and higher “knee” temperature than Ru(Cp)2. However, at lower temperatures (T<320° C.), the wafers deposited with Ru(Cp)2 form primarily the RuO2 phase. Thus, below 320° C., the activation energy is mainly controlled by the formation of RuO2 rather than ruthenium making it difficult to compare the activation energies for Ru(EtCp)2 and Ru(Cp)2.

[0086] The plots of resistivity temperature dependence show resistivity increasing as temperature decreases for both precursors. The increase in resistivity is more dramatic for Ru(Cp)2. The higher resistivity for Ru(Cp)2 is probably due to the formation of RuO2, which has a higher bulk resistivity (45&mgr;&OHgr;-cm) than pure Ru (7&mgr;&OHgr;-cm).

EXAMPLE 7

[0087] Ruthenium Film Properties

[0088] The CVD Ruthenium film properties and process performance are summarized below in Table 3. 3 TABLE 3 Comparison of CVD Ru film properties and process performance with RU(Cp)2 in tetrahydrofuran and Ru(EtCp)2 in octane at different process temperatures on different substrates. &rgr; WTW WIW WIW XRD Haze (ppm)/ D/R thk thk &rgr; % 1&sgr; Int. Roughness (Å Step Data Process Å/min % 1&sgr; %1&sgr; &mgr;&OHgr;-cm 5 mm Ord. RMS, Rmax) Cov/Conf. % Ref. Ru(Cp)2 70  1.8% 2-6% by 20 31%  97 ppm/31.4Å S/T:40 100-water 340° C. Meta  305Å B/T:30 run Ru91108 on pulse (470Å flim) C:60 TiN/ThOx Ru(Cp)2 70 N/A N/A 15  5% 002 112 ppm/62.2Å S/T:30 Q9K11442 340° C. 101  491Å B/T:35 on 100 (710Å film) C:60 PVD Ru/ TiN/ Si Ru(Cp)2 50 10.2% N/A N/A S/T:90 Ru91029 320° C. B/T:90 on TiN C:80 Ru(Cp)2 60 N/A N/A 38 20% 101 230 ppm/59.0Å Q9K11438 320° C. 002  492Å on 100  588Å film PVD Ru/ TiN/ Si Ru(EtCp)2 160 N/A N/A 35 60% 730 ppm/122Å 91103_5 350° C.  954Å on TiN 1430Å film Ru(EtCp)2 10 N/A N/A N/A N/A N/A N/A N/A 330°C on TiN Ru(EtCp)2 120  1.2% TBD 42  8% 100  39 ppm/663Å S/T:90 25-wafer 330° C. 002  541Å B/T:90 run on PVD 101 (700Å film) C:80 91119 Ru Abbreviations: S/T is sidewall thickness divided by top thickness. B/T is bottom thickness divided by top thickness. C is conformality which is defined as the thinnest part of the test structure sidewall divided by the thickest part of the sidewall

EXAMPLE 8

[0089] Kinetic Study Using Pure Ru(EtCp)2

[0090] Using a pure Ru(EtCp)2 precursor lowers the activation energy to 0.8-0.9 eV. Activation energy is also dependent on the rate of Ru flow. Using a CVD Arrhenius plot, comparing the Ru flow rates of 20, 50 and 80 mgm with concomitant O2 flows of 40, 160 and 100 sccm and a constant N2 flow on deposition rates on a 50 Å PVD Ru/2 kÅ ThOx substrate indicates that best conditions occur at Ru flow 50 mgm with an activation energy of 0.79 eV (FIG. 5). The higher activation energies at low temperatures for high Ru flow may possibly be due to a lack of N2 flow. At an Ru flow rate of 50 mgm, uniformity increases rapidly as the temperature decreases to the optimal 335° C. Under these conditions Ru CVD yields a silver Ru film with a resistivity of 22&mgr;&OHgr;-cm (FIG. 6A). At an Ru flow of 20 mgm on the same substrate, uniformity increases dramatically below 345° C., a result of increased temperature sensitivity, and resistivity remains fairly constant at temperatures below 335° C., (FIG. 6B); under these process conditions the heater has a 4° C. max.-min. variation for setpoints from 300-350° C.).

[0091] Increasing the N2 flow increases the deposition rate of Ru films on a 50 Å PVD Ru/2 kÅ ThOx substrate and lowers the activation energy in the kinetic-limited regime (FIG. 7), whereas a higher N2 flow reduces the deposition rate in the mass transport regime. Additionally, at high N2 flow, 1300 sccm, and 50 mgm Ru flow on the same substrate, resistivity increases at lower deposition temperatures (FIG. 8). Generally, increasing the N2 flow rate improves the Ru CVD process, making it more stable and less temperature dependent. It is therefore possible to tune the nitrogen and oxygen flow rates and deposition temperature to achieve the desired Ru film.

EXAMPLE 9

[0092] X-Ray Diffraction Data of RU(Cp)2 (0.1 M in tetrahydrofuran) and Ru(EtCp)2 (1.0 M in octane)

[0093] FIG. 9A shows an X-Ray diffraction plot of films grown with Ru(EtCp)2 precursor in octane with a deposition temperature of 330° C., in the kinetic-limited regime. This film has a relatively strong Ru(100) orientation as well as Ru(002) and Ru(101) peaks. Although no peaks for RuO2 were present, the resistivity of this film is greater than 30&mgr;&OHgr;-cm and RBS analysis shows that the film has 30±10% oxygen. This film was shiny and metallic in appearance. At 320° C., the X-Ray diffraction pattern (FIG. 9B) shows some evidence of RuO2 formation with weak RuO2(110) and RuO2(002) peaks, but the dominant phase was Ru.

[0094] CVD Ru films grown with the Ru(Cp)2 precursor in tetrahydrofuran have different crystalline orientation than films grown with Ru(EtCp)2. FIG. 9C shows an X-Ray Diffraction pattern of CVD Ru grown with Ru(Cp)2 at 340° C., in the mass-transfer limited regime. This film has a relatively strong Ru(002) peak as well as Ru(101) and Ru(100) peaks. At 320° C., some RuO2 phase was mixed with the Ru phases (FIG. 9D). In the kinetic-limited regime at 295° C., the film was almost completely RuO2 (FIG. 9E).

[0095] The extent of oxidation of the films grown with Ru(Cp)2 in the kinetic-limited temperature regime appears to be greater than the oxidation of films grown with Ru(EtCp)2. In fact, it has not been possible to grow films with mostly pure Ru phase at low temperatures using Ru(Cp)2 using a wide range of gas flows and pressures. On the other hand, the O2 flow can be varied using the, Ru(EtCp)2 precursor to form mostly pure Ru phase or mostly RuO2 phase. In this way the Ru(EtCp)2 precursor has the advantage of tuning the film Ru/O composition ratio under low-temperature conditions where conformal films can be deposited.

[0096] The orientation of CVD Ru films depends on the substrate even when a PVD Ru, seed is used. FIG. 10 shows the X-Ray Diffraction patterns of CVD Ru using Ru(EtCp)2 on two slightly different substrates each having a ˜60 Å PVD Ru seed layer. One wafer had 2000 Å thermal oxide under the PVD Ru layer while the other had 200 Å PVD TiN. The CVD Ru film with the Ru seed on thermal oxide has a very strong Ru(101) peak while the film with Ru seed on TiN had only weak Ru peaks. This result emphasizes the significance of the substrate on the film properties even when a thin PVD Ru seed layer was used.

EXAMPLE 10

[0097] X-Ray Diffraction Data of Ru Films Grown with Pure Ru(EtCp)2

[0098] As with the Ru(EtCp)2 precursor in octane, pure Ru(EtCp)2 precursor can be used to tune the Ru/O film composition ratio to control the crystalline orientation of the Ru films being deposited. The best Ru(002) orientation occurs at the lowest O2 flow, FIG. 11 shows the effect of O2 flow on CVD Ru orientation on a 50 Å PVD Ru/2 kÅ ThOx substrate at 355° C. RuO2 formation begins to occur at O2 flow 1200 sccm as evidenced by the abrupt appearance of RuO2 (101) peaks and, to a lesser extent, RuO2 peaks at 200 and 110.

[0099] An examination of the effect of O2 flow on CVD Ru (002) orientation at 355° C. using process conditions of Ru 50 mgm, N2 250 sccm and 8 torr of pressure, again indicates an abrupt increase in the relative intensity of the 002 peak at O2 flow 1200 sccm (FIG. 12). The concomitant increase in resistivity to greater than 30&mgr;&OHgr;-cm suggests the appearance of RuO2 in the deposited film. Below ˜900 sccm O2 flow, the film is silver and has a high Ru content which becomes a more brown/purple hue as the RuO2 content increases with an increase in the O2 flow rate.

EXAMPLE 11

[0100] Within-Wafer Film Thickness Uniformity

[0101] Measuring the within-wafer thickness uniformity of CVD Ru by XRF has been limited due to a background signal that varies from the center to the edge of the wafer. Uncorrected WIW thickness uniformity values by XRF are typically 10-15% 1&sgr;. The WIW thickness uniformity for blank substrates (which should measure zero thickness everywhere) is also very high. It is necessary to zero the background signal at all points on the wafer surface to ensure that reliable WIW measurements are made.

[0102] The Rudolph Metapulse is another tool that may be capable of measuring within-wafer Ru thickness. Initial data using the Ru(Cp)2 precursor on TiN/Si substrates indicates the WIW thickness uniformity is as low as 2% 1&sgr;. However, a discrepancy exists between the WIW thickness uniformity numbers and the XRF measurement. Possibly the Metapulse measurement is affected by poor adhesion between the CVD Ru -and TiN layers. A second set of CVD Ru films grown on 60 Å. PVD Ru/2 kÅ SiO2 yielded very weak signals. In this instance the oxide thickness could be reducing the film reflectivity. Metapulse measurements using SiO2 substrates less than 1000 Å thick may correct the problem.

EXAMPLE 12

[0103] Within-Wafer Sheet Resistance Uniformity

[0104] Sheet resistance was measured by the KLA-Tencor 4-point probe analyzer. The within-wafer sheet resistance uniformity was taken from 49 points measured across the wafer using a 5 mm edge exclusion. A 5 mm edge exclusion was used to accommodate the PVD seed layer or adhesion layer which has a 3 mm edge exclusion from the clamp ring. The MRS requirement for WIW Rs uniformity (49 pt) was 10% 1&sgr;. The CVD Ru films using either precursor were able to achieve WIW Rs uniformity below 10% 1&sgr; using the PVD Ru seed layer. Table 3 above shows the dramatic improvement in Rs uniformity by using a PVD Ru seed layer. For Ru(Cp)2, the Rs uniformity was reduced from 30% to 5% 1&sgr;.

EXAMPLE 13

[0105] Dependence of Sheet Resistance on Film Thickness and Deposition Temperature

[0106] The CVD Ru film resistivity (&rgr;) increases with decreasing deposition temperature. Typical &rgr; vs T data is shown on the Arrhenius plots for the Ru(Cp)2 (FIG. 4B) and Ru(EtCp)2 (FIG. 4D) precursors. The trend was not linear but was instead sigmoidal; a sharp increase in film resistivity is seen near the transition from mass-transfer to kinetic-limited temperature regime. This increase in resistivity was most likely the result of increasing O incorporation in the Ru film with decreasing deposition temperatures. XRD patterns confirm the emergence of RuO2 peaks at lower process temperatures. The extent of oxidation in the kinetic-limited temperature regime is greater for films grown with the Ru(Cp)2 precursor relative to the Ru(EtCp)2 in octane precursor, as evidenced by greater intensities of RuO2 peaks and higher film resistivity (FIG. 9E).

EXAMPLE 14

[0107] Surface Roughness Comparison

[0108] The surface roughness and morphology of Ru films deposited by CVD varies depending on the wafer temperature and substrate. CVD Ru films were deposited on PVD TiN as well as on PVD Ru seed layers with PYD TiN. SEM images were taken of the ruthenium films to compare the surface morphologies. CVD Ru films deposited directly onto PVD TiN using Ru(EtCp)2 in octane precursor had very rough surfaces (FIG. 13A). Using the same process conditions, the surface morphology was improved by using a thin PVD Ru seed layer, approximately 60 Å in thickness below the CVD Ru film (FIG. 13B). Both films were deposited using Ru(EtCp)2 in octane precursor at a wafer temperature of 343° C.

[0109] With the Ru(Cp)2 precursor, the surface roughness of the CVD Ru film was slightly better on PVD TiN than on the PVD Ru seed. FIG. 14 shows the AFM 2 &mgr;m×2 &mgr;m scans comparing CVD Ru on the PVD Ru seed with the PVD TiN substrate. Both films were deposited using the same 343° C. Ru(Cp)2 process. Even though the RMS values cannot be directly compared because CVD Ru thickness was 50% greater for the sample on PVD Ru than the one on PVD TiN, the RMS roughness value was still 100% greater.

[0110] Comparing the surface morphology at 343° C. using a PVD Ru seed layer, the film deposited with Ru(Cp)2 looks smoother than the films deposited with Ru(EtCp)2 (FIGS. 13C and 13D). At a wafer temperature of 343° C., the process was mass-transfer limited for both precursors. In Table 4, AFM results show the roughness values to be fairly high for both films. The RMS value is 121.9 Å for a 1430 Å film deposited with the Ru(EtCp)2 precursor and 62 Å for a 710 Å film deposited with the Ru(Cp)2 precursor.

[0111] FIGS. 13E and 13F compare the two films at their respective temperatures in the kinetic-limited regime. At a wafer temperature of 320° C., the film deposited with Ru(Cp)2 has raised and depressed regions which are detectable at lower magnifications, 2,000 to 10,000 times. This overlying macroscopic pattern causes visible haze on the wafer. FIG. 13F shows that the film deposited using Ru(EtCp)2 at 331° C. had small grains with sharp, high aspect ratio features but no overlying pattern. The RMS value for a 540 Å film deposited with Ru(EtCp)2 was 66 Å. 4 TABLE 4 AFM Roughness Values for CVD Ru Films Thickness RMS Ra Rmax Film Description (Å) (Å) (Å) (Å) 1. Ru(EtCp)2 on PVD  540 Å  66.3 Å 51.4 Å 540.6 Å Ru seed at 331° C. 2. Ru(EtCp)2 on PVD 1430 Å 121.9 Å 97.5 Å 954.4 Å Ru seed at 343° C. 3. Ru(Cp)2 on PYD  588 Å  59.0 Å 46.9 Å 491.5 Å Ru seed at 320° C. 4. Ru(Cp)2 on PVD  710 Å  62.0 Å 49.7 Å 490.8 Å Ru seed at 343° C. 5. Ru(Cp)2 on PVD  470 Å  31.4 Å 24.9 Å 305.0 Å TiN at 343° C.

[0112] Overall, the CVI) Ru films deposited using either Ru(EtCp)2 or Ru(Cp)2 as precursor appear to be quite rough. The average roughness RMS values were about 10% of the total film thickness for each of the conditions. The surface roughness does not seem to change dramatically with temperature. However, surface morphology was affected by the wafer temperature and the grain size tends to increase with temperature. For Ru(EtCp)2 precursor, the surface roughness improves by using a PVD Ru seed layer relative to a PVD TiN layer. The opposite trend was observed for the Ru(Cp)2 precursor.

[0113] Although surface roughness is an important consideration with the CVD Ru films, the actual roughness requirement has not been defined. It is not clear which precursor, Ru(EtCp)2 or Ru(Cp)2, produces smoother films. The substrate seems to be a strong variable that affects surface roughness and morphology. Additionally, process parameters such as pressure, carrier gas flow, and spacing, also can affect surface roughness.

EXAMPLE 15

[0114] Step Coverage

[0115] Ru(Cp)2 and Ru(EtCp)2 in octane precursor performance for step coverage and conformality were compared (FIGS. 15-20). The comparisons include a) deposition temperature, b) type of Ru precursor, and c) substrate (PVD Ru seed layer vs. TiN). Conformality is defined as the thinnest part of the test structure sidewall divided by the thickest part of the sidewall. Step coverage refers to sidewall (sidewall thickness/top thickness) and bottom coverage (bottom thickness/top thickness). Average thicknesses were used for the step coverage calculation.

[0116] For CVD processes, lower temperatures in the kinetic-limited regime were typically required to achieve good step coverage and conformality. FIG. 15 shows films deposited on PVD TiN using Ru(Cp)2 at two different wafer temperatures, 343° C. (FIG. 15A) and 320° C. (FIG. 15B). At 343° C. the step coverage for a 0.17 &mgr;m structure with AR 5:1 is ˜40% on the sidewall and ˜30% on the bottom while the conformality is ˜60%. The SEM cross-section images show better step coverage for the film deposited at 320° C., below the “knee” in the kinetic-limited regime. Step coverage (sidewall and bottom) for the 320° C. process is ˜90%, and conformality is ˜80% for a 0.17 &mgr;m structure at an aspect ratio (AR) of 5:1. However, an SEM of a structure with larger geometry (0.28 &mgr;m AR 3:1) reveals that the film is very rough and not continuous in some areas.

[0117] FIG. 16 compares the step coverage of Ru(Cp)2 (FIG. 16A) and Ru(EtCp)2 in octane (FIG. 16B) films at 343° C. on PVD TiN substrates. The SEM images reveal that the Ru(EtCp)2 in octane film at 340° C. is extremely rough with faceted and discontinuous grains.

[0118] By depositing CVD Ru onto a PVD Ru seed layer, the film roughness and step coverage can be greatly reduced. FIG. 17 compares CVD Ru films with (FIG. 17B) and without (FIG. 17A) a PVD Ru seed layer using Ru(EtCp)2 in octane. Because the CVD Ru process is slightly substrate dependent, the film without the PVD Ru seed layer is deposited at 340° C.; however, the two films are deposited in similar temperature regimes. The conformality and step coverage were much better with the PVD Ru seed layer, and the surface was smoother. The PVD Ru seed layer is a good template substrate for CVD Ru growth on patterned structures.

[0119] FIG. 18 shows SEM crosssections of CVD Ru films deposited onto ˜60 Å PVD Ru seed layers. The film deposited using the Ru(EtCp)2 in octane (FIG. 18B) precursor demonstrates much better step coverage with ˜90% sidewall and bottom coverage for a feature size, of 0.15 &mgr;m and AR 6:1. The Ru(Cp)2 (FIG. 18A) film only showed 50% sidewall coverage and 40% bottom coverage for the same structure. Both films have about 80% conformality that is higher than the films deposited without the PVD seed layer.

[0120] FIG. 19 shows an SEM crosssection of a CVD Ru film using pure Ru(EtCp)2 on a 50 Å PYD Ru/2 k Å ThOx substrate at 355° C. for a feature size of 0.15 &mgr;m and AR 6:1. Rs film uniformity is ˜8% under these conditions. Top film thickness is 400 Å and side film thickness is ˜350 Å with both sidewall and bottom coverage at ˜88%.

[0121] FIG. 20 shows the step coverage and conformality of CVD Ru deposited directly on a Ta2O5/poly-Si substrate without a PVD Ru seed layer at 340° C. The CVD Ru can not be deposited at the best known process temperature of 330° C. used for PVD Ru seed layers. The step coverage and conformality were quite good considering the narrow spacing and severe aspect ratios. The CVD Ru surface was smooth, even in the absence of PVD Ru seed, Thus, CVD Ru can be used as a top electrode for TanOx MIM applications even for very high aspect ratios.

[0122] The SEM images in FIG. 21 indicate that CVD Ru films do not grow uniformly on CVD BST films. CVD Ru was deposited on 300 Å BST/PVD Pt on 0.2 &mgr;m wide cup structures with AR 1.5:1. As with the deposition on Ta2O5, the process temperature had to be increased to 340° C. From the SEM photos, it appears that the CVD Ru growth is not controlled well on the BST surface. Large crystallites of ruthenium are seen randomly spread across the wafer inside and outside of the cup structures. In some areas, the ruthenium seems to be depositing along the sidewalls but the film looks rough.

[0123] In summary, conformality and step coverage improve at lower temperatures, especially in the kinetic-limited temperature regime. Lower deposition temperatures also reduce surface roughness. Both Ru(Cp)2 and Ru(EtCp)2 produce CVD Ru films with good step coverage and conformality in the kinetic-limited temperature regime; however, the deposition rate is higher using Ru(EtCp)2 in comparable temperature regimes resulting in greater wafer throughput. Additionally, using a PVD Ru seed layer reduces surface roughness and appears to improve nucleation of CVD Ru films deposited with Ru(EtCp)2.

EXAMPLE 16

[0124] Process Condition Repeatability

[0125] The wafer-to-wafer thickness for both the Ru(Cp)2 (FIG. 22A) and Ru(EtCp)2 (FIG. 22B) precursors meet the C&F goal of less than 2% 1 sigma. However, the run with Ru(EtCp)2 has a process drift, with the average thickness increasing 10-15 Å over 25 wafers. As this is only a 25 wafer run, longer run is needed to determine whether this trend of increasing thickness is an anomaly.

EXAMPLE 17

[0126] Edge Exclusion Ring

[0127] An aluminum clamp ring for edge exclusion of CVD Ru deposition was tested on a total of 8 wafers using the BKM CVD Ru process. Edge exclusion of CVD Ru is a common requirement and may be accomplished by means other than a clamp ring such as is well known by those having ordinary skill in this art. It is noted that the clamp ring was not centered on the wafer (FIG. 23A). This pattern was present for all 8 wafers.- A centering mechanism would make the clamp ring concentric to the wafer. The film edge was reasonably well-defined at the top half of the wafer, suggesting a good contact to the wafer surface (FIG. 23B). The edge exclusion was primarily opposite the notch and extended approximately over the top half of the wafer only, as indicated by optical microscope observations of the film edge (FIG. 23C). On the very first wafer, particles were observed under the clamp ring at the top half of the wafer (FIG. 23D). The number of particles was reduced after several wafers were run.

[0128] The effects of the clamp ring on the CVD Ru process are identified below in TABLE 5. The clamp ring was allowed to touch the wafer for 60 seconds before the start of the deposition. Heater-showerhead spacing was 400 mil. 5 TABLE 5 WIW thk Rs Deposition uniformity Film uniformity Avg Rate % 1&sgr;) Resistivity (%1&sgr;) Haze (Å/min) 15 mm e.e. (&mgr;&OHgr;-cm) 5 mm e.e. (ppm) Without 123 Å/min 15%* 50  7% 30 clamp ring (5 mm e.e.) With 118 Å/min 13%* 60 30% 30 clamp (5 mm e.e.) ring 13% (15 mm e.e.)

[0129] data taken with Ru(EtCp)2 precursor at 330° C. using PVD Ru seed/TiN/Si substrates; data with clamp ring Ru91117; data without clamp ring Ru91124. *Note: high WIW uniformity due to varying background signal (wafer center to edge) in XRF. WIW uniformity is typically less than 5% 1_ from Metapulse (not measured on both sets of samples).

[0130] The biggest impact of the clamp ring was on the within-wafer Rs uniformity. The clamp ring may be affecting the wafer temperature near the edge and thus alter the film, The WIW thickness uniformity was similar; however, the current XRF measurement for WIW uniformity was not accurate. In addition, the deposition rate was slightly lower, and the film resistivity was slightly higher.

EXAMPLE 18

[0131] Metrology

[0132] One possible method of measuring CVD Ru thickness is using the sheet resistance measurement of the 4-point probe. FIG. 24 shows how sheet resistance varies with film thickness. If the film resistivity was assumed to be constant over this thickness range, then this is essentially a 1/thickness vs. thickness curve. With a proper calibration, the film thickness could be determined from the measured sheet resistance.

[0133] The within-wafer thickness uniformity of CVD Ru films was difficult to measure by XRF due to a large background signal from the silicon substrate that varies from wafer center to edge. While it may be possible to correct for this problem using proper calibration procedures, an alternative method is the Rudolph Metapulse tool, which measures laser-induced sound wave pulses in the film to obtain film thickness.

EXAMPLE 19

[0134] Ru(Cp)2 Vs. Ru(EtCp)2 Precursors: Summary

[0135] The Ru(Cp)2 precursor has a limited solubility (0.12M in THF) which limits the CVD deposition rate. The current deposition rate for the 343° C. process using Ru(CP)2 is about 70 Å/min in the mass-transfer-limited regime. In contrast, the Ru(EtCp)2 precursor, which is a liquid at room temperature and is currently used at 1M in octane, has a current deposition rate of greater than 300 Å/min in the mass-transfer-limited regime and greater than 100 Å/min for the BKM process in the kinetic-limited regime.

[0136] The residue from the TGA curve is zero for Ru(EtCp)2, while it is about 2% for Ru(Cp)2. Finite residues are not desirable because residues can result in particles and/or process drift.

[0137] The step coverage using the Ru(EtCp)2 precursor with a PVD Ru seed layer was close to 90% with the BKM process. The step coverage on TanOx substrates using Ru(EtCp)2 for top electrode applications has excellent filling capabilities to greater than 10:1 AR.

[0138] The wafer-to-wafer thickness repeatability using Ru(EtCp)2 meets the MRS alpha-exit requirements of less than 2% 1 &sgr;.

[0139] Compared with Ru(Cp)2, Ru(EtCp)2 has a wider process window for tuning the O/Ru film ratio. Ru(EtCp)2 is less susceptible to oxidation during MOCVD Ru deposition at low temperatures. A low-oxygen Ru film with low film resistivity can be formed in the kinetic-limited temperature regime using certain process conditions. By changing process parameters, e.g., higher O2 flow, it is also possible to form high-resistivity RuO2 films with Ru(EtCp)2. This is in contrast with Ru(Cp)2 which has a greater tendency to form high-resistivity RuO2 films in the kinetic-limited regime for nearly all process conditions. See Table 1 for a summary of CVD Ru film deposition using Ru(EtCp)2 as the precursor as compared with Alpha Exit Specifications. Table 6 provides a summary of the processing parameter ranges for forming an Ru film using using Ru(EtCp)2 as the precursor for metalorganic chemical vapor deposition (MOCVD) while Table 7 provides specific process parameters for optimal deposition of Ru films. 6 TABLE 6 Processing Parameter Ranges for Forming a Ru Film Process Parameter. Range Carrier Gas Argon, Helium, Xenon, Neon; Krypton, Nitrogen Carrier Gas Flow in Vaporizer 100-1000 sccm for Ru Precursor (sccm = standard cubic centimeters per minute) Push gas pressure  20-200 psi Oxygen Flow 100-3000 sccm Nitrogen Flow (to chamber) 100-3000 sccm Ru precursor Ru(Cp)2, Ru(EtCp)2, Ru(iPrCp)2, Ru(MeCp)2, Ru(thd)3, Ru(OD)3 Me = methyl, iPr = isopropyl, Cp = cyclopentadienyl, Et = ethyl, thd = tetramethylheptanedionate, OD = octanedionate Ru Concentration 0.01M - neat (pure) Solvent (can be pure or 10. Alkanes (e.g., octane, heptane, decane, hexanes, etc.) mixtures of these) 11. Aromatic hydrocarbons (benzene, toluene, xylenes, etc.) 12. Ethers(diethyl ether, dimethyl ether, etc.) 13. Cyclic ethers (tetrahydrofuran, tetrahydropyran, etc.) Solvent additives Stabilizers/complexing agents such as EDTA, ethlenediamine, pentamethyl-diethylenetriamine, etc. Ru flow   10-500 mg/min Ru vaporizer temperature  100-300° C. Jackets/Lid temperature  100-300° C. Feedthrough temperature  100-300° C. Substrate support member  150-500° C. temperature Substrate Temperature  100-500° C. Showerhead/substrate support  100-800 mils member spacing Chamber pressure  0.1-100 torr

[0140] 7 TABLE 7 Example Parameters for Forming a Ru Film Process Parameter Range Carrier Gas Nitrogen Carrier Gas Flow in Vaporizer for Ru 450 sccm Precursor Push gas pressure  80 psi Oxygen Flow 300 sccm Nitrogen Flow 250 sccm (to chamber) Ru precursor Ru(EtCp)2 Ru Concentration 1 M in octane (29% by weight) Ru flow 180 mg/mm Ru vaporizer temperature 260° C. Jackets/Lid temperature 260° C. Feedthrough temperature 260° C. Substrate support member 345° C. temperature Substrate Temperature 340° C. Showerhead/substrate support 350 mils member spacing Chamber pressure 8 torr

[0141] Any patents or publications mentioned in this specification are indicative of the levels of those skilled in the art to which the invention pertains. These patents and publications are herein incorporated by reference to the same extent as if each individual publication was specifically and individually indicated to be incorporated by reference.

[0142] One skilled in the art will readily appreciate that the present invention is well adapted to carry out the objects and obtain the ends and advantages mentioned, as well as those inherent therein. It will be apparent to those skilled in the art that various modifications and variations can be made in practicing the present invention without departing from the spirit or scope of the invention. Changes therein and other uses will occur to those skilled in the art which are encompassed within the spirit of the invention as defined by the scope of the claims.

Claims

13. (New) A method of depositing a ruthenium-containing film on a substrate comprising:

vaporizing a neat Ruthenium liquid source material in a flow of a carrier gas;

exposing the vaporized ruthenium source material to a flow of oxygen gas;

reacting the vaporized ruthenium source material with the oxygen at an activation energy (Ea) of less than 1 eV in a kinetically limited temperature regime to deposit a Ruthenium-containing film on the substrate.

14. (New) The method of claim 13 wherein the deposited Ruthenium-containing film comprises Ruthenium metal.

15. (New) The method of claim 13 wherein the deposited Ruthenium-containing film comprises Ruthenium oxide.

16. (New) The method of claim 13 wherein the Ruthenium-containing film is formed in the absence of a seed layer overlying the substrate.

17. (New) The method of claim 13 further comprising forming a seed layer on top of the substrate prior to deposition of the Ruthenium-containing film.

18. (New) The method of claim 17 wherein said seed layer is formed by a vapor deposition technique selected from the group consisting of physical vapor deposition and chemical vapor deposition.

19. (New) The method of claim 18 wherein said seed layer is selected from the group consisting of ruthenium, iridium, platinum, titanium nitride, titanium aluminum nitride, tantalum pentoxide, ruthenium oxide, iridium oxide, and titanium silicide.

20. (New) The method of claim 13 wherein said neat Ruthenium liquid source material is bis-(ethylcyclopentadienyl) ruthenium.

21. (New) The method of claim 13 wherein the neat Ruthenium liquid source material is vaporized at a temperature of about 260° C.

22. (New) The method of claim 13 wherein the Ruthenium liquid source material is vaporized in a flow of an N2 carrier gas.

23. (New) A method of depositing a ruthenium-containing film on a substrate comprising:

vaporizing a Ruthenium liquid source material in a flow of a carrier gas at a first flow rate;

exposing the vaporized ruthenium source material to a flow of oxygen gas at a second flow rate;

reacting the vaporized ruthenium source material with the oxygen gas an elevated temperature in a kinetically limited temperature regime to deposit a Ruthenium-containing film on the substrate; and

controlling at least one process parameter selected from the group consisting of a concentration of the Ruthenium liquid source material, a vaporization temperature of the Ruthenium liquid source, a deposition temperature, the first flow rate, and the second flow rate to determine a property of the deposited Ruthenium-containing film.

24. (New) The method of claim 23 wherein the deposited Ruthenium-containing film comprises Ruthenium metal.

25. (New) The method of claim 23 wherein the deposited Ruthenium-containing film comprises Ruthenium oxide.

27. (New) The method of claim 23 wherein vaporizing a Ruthenium liquid source material comprises vaporizing a neat Ruthenium liquid source material.

28. (New) The method of claim 27 wherein said neat Ruthenium liquid source material is bis-(ethylcyclopentadienyl) ruthenium.

29. (New) The method of claim 23 wherein vaporizing a Ruthenium liquid source material comprises vaporizing a Ruthenium liquid source material dissolved in a solvent.

30. (New) The method of claim 29 wherein the Ruthenium liquid source material is bis-(ethylcyclopentadienyl) ruthenium and the solvent is selected from the group consisting of octane and THF.

31. (New) The method of claim 23 wherein the first flow rate is controlled to determine a rate of deposition of the Ruthenium-containing film.

32. (New) The method of claim 23 wherein the vaporization temperature is controlled to determine a rate of deposition of the Ruthenium-containing film.

33. (New) The method of claim 23 wherein the second flow rate is controlled to determine a Ru/O composition ratio of the Ruthenium-containing film.

34. (New) The method of claim 23 wherein the second flow rate is controlled to determine crystal alignment of the Ruthenium-containing film.

35. (New) The method of claim 23 wherein the deposition temperature is controlled to determine a Ru/O composition ratio of the Ruthenium-containing film.

36. (New) The method of claim 23 wherein the concentration of the Ruthenium liquid source material is controlled to determine a Ru/O composition ratio of the Ruthenium-containing film.

37. (New) The method of claim 23 further comprising forming a seed layer on top of the substrate prior to deposition of the Ruthenium-containing film.

38. (New) The method of claim 37 wherein said seed layer is formed by a vapor deposition technique selected from the group consisting of physical vapor deposition and chemical vapor deposition.

39. (New) The method of claim 37 wherein said seed layer is selected from the group consisting of ruthenium, iridium, platinum, titanium nitride, titanium aluminum nitride, tantalum pentoxide, ruthenium oxide, iridium oxide, and titanium silicide.