摘要:

We have completed a comparison of SiO2film thicknesses obtained with the three dominant measurement techniques used in the Integrated Circuit industry: ellipsometry, capacitance‐voltage (C‐V) measurements and high resolution transmisission electron microscopy (HRTEM). This work is directed at evaluating metrology capability that might support NIST‐ traceable Reference Materials for very thin dielectric films. Particular care was taken in the design of the sample set to allow redundancy and enable estimates of oxide layer consistency. Ellipsometry measurements were analyzed using a variety of models of the film structure, and C‐V results were analyzed using three different quantum‐mechanical based algorithms to account for quantized states in the substrate and depletion effects in the polysilicon capacitor electrode. HRTEM results were supplemented with Electron Energy‐Loss Spectroscopy. A range of thicknesses was found with each of the methods, but with some overlap of values. HRTEM and STEM values showed less consistency between wafers than the C‐V data for the capacitors used and were seen to be more influenced by local variations such as interface non‐uniformities. Sources of variation and estimates of uncertainty for the analyses are presented. Implications of these results for Reference Materials are discussed. © 2003 American Institute of Physics

ISSN:0094-243X    

DOI:10.1063/1.1622491

出版商:AIP    

年代:1903

数据来源: AIP

摘要:

Grazing incidence X‐ray reflectivity (GIXR), ellipsometry and X‐ray photoelectron spectroscopy (XPS) have been used for measuring the thickness of ultrathin SiO2films on Si(100). SiO2films were fabricated at a constant temperature to obtain a fixed interface structure for films with different thicknesses. The thicknesses obtained by GIXR and ellipsometry were in good agreement with each other, however, ellipsometry showed slightly larger values. The thickness of the adsorbed overlayer was also compared using GIXR and XPS. Uncertainties included in the XPS measurements of the carbonaceous layer thickness were estimated. The thicknesses of the carbonaceous layer obtained by XPS were slightly smaller, by about 0.16 nm, than those of the adsorbed overlayer obtained by GIXR. About 0.3–0.4 monolayer of adsorbed water molecules is believed to account for the differences in overlayer thicknesses between the GIXR and XPS measurments. © 2003 American Institute of Physics

ISSN:0094-243X    

DOI:10.1063/1.1622492

出版商:AIP    

年代:1903

数据来源: AIP

摘要:

A modified Ultrasonic Force Microscopy (UFM) technique is presented that determines the elasticity (Young’s modulus) of a material quantitatively based on the Johnson‐Kendall‐Roberts model. The periodic oscillation of the sample with MHz frequencies causes a high dynamical stiffness of the cantilever of a scanning probe microscope (SPM). The information about the elasticity of a sample can be extracted from the cantilever response to the amplitude variation. The procedure for the quantitative determination of the reduced Young’s modulus is demonstrated for a partially delaminated thin film. For such a sample, no material inhomogeneities exist for the top layer, i. e., surface roughness and sample‐tip adhesion are assumed to be constant. Consequently, the UFM signal is not influenced by locally fluctuating material properties and topography. The systematic study shows that film delamination causes a continuous gradient of film stiffness along the delaminated part of the film. The application of this modified UFM technique in physical failure analysis is demonstrated for the nondestructive characterization of buried defects. © 2003 American Institute of Physics

ISSN:0094-243X    

DOI:10.1063/1.1622493

出版商:AIP    

年代:1903

数据来源: AIP

摘要:

Because high‐resolution transmission electron microscopy (HRTEM) relies on a complex contrast mechanism to produce images of gate dielectric films in cross section, there are many factors affecting the uncertainty of thickness measurements based on these images. A preliminary survey revealed approximately 50 parameters that affect the uncertainty in a gate dielectric dimensional metrology experiment using HRTEM, along with approximately 1,200 two‐term interactions and almost 20,000 three‐term interactions. Using established design‐of‐experiment (DEX) methodologies, I performed a screening experiment based on a2IV(8−4)fractional factorial design to determine which factors had the greatest impact on the absolute error of the thickness measurements. Absolute error was determined by simulating HRTEM micrographs using a multislice calculation. The model used for the simulation consisted of a variable SiO2film approximately 2 nm thick positioned between two pieces of crystalline Si. This approximation to a gate stack was built atom‐by‐atom using commercial molecular modeling software supplemented with custom Tcl scripts to assemble the gate structures from simpler primitives. By varying the molecular model, sample parameters such as crystallographic orientation, film thickness, density, and along‐beam thickness can be varied precisely. Instrument parameters and details of the imaging conditions are inputs to the multislice calculation, a simulation technique that has been vetted by the microscopy community and has been in use for decades. Beam tilt, defocus, and vibration amplitude were the main factors found to have the largest effects, while beam‐tilt↔defocus and defocus↔vibration were the most important two‐term interactions. © 2003 American Institute of Physics

ISSN:0094-243X    

DOI:10.1063/1.1622494

出版商:AIP    

年代:1903

数据来源: AIP

摘要:

Because thin films are formed by processes different from those used to produce bulk materials, their microstructures, and hence their mechanical properties, are quite different from those of bulk materials of the same chemical composition. While the general principles of conventional mechanical testing are applicable to thin films, special test equipment and techniques are required. These are briefly described here. Present specimen sizes are near the limit of what can be tested in the optical microscope, so techniques useful in the scanning electron microscope are of interest. Test techniques adapted for use in the SEM are presented. These test methods have been applied to pure aluminum films deposited in our laboratory, aluminum films made in a commercial CMOS fab facility, electrodeposited copper, polyimide films, and polysilicon films. The differences among the stress‐strain curves for these very different materials were as dramatic as would be expected. Now that some experience with these test techniques has been accumulated and the reproducible results are becoming available, comparisons can be made to expectations based on well‐established bulk behavior. Current unresolved materials‐science issues include the “deficit” of the quasi‐static apparent Young’s modulus relative to bulk values of some metals, and the generally low elongation to failure found in tensile tests of free‐standing metal films. © 2003 American Institute of Physics

ISSN:0094-243X    

DOI:10.1063/1.1622495

出版商:AIP    

年代:1903

数据来源: AIP

摘要:

A novel non‐destructive method for volumetric thickness profile measurement of a transparent film, thinner than the white light coherence length of 3∼4 &mgr;m, that is deposited on a patterned structure is described in this paper. A visible acousto‐optic tunable filter (AOTF) for wavelength scanning is employed and the 3 dimensional thin film thickness profile information is obtained through 2 phase functions &psgr;(k)and &fgr;(k)in the spectral domain. The first phase function &psgr;(k)compensates for the phase change effect caused by the multiple reflected beams from the thin film and the other is a total phase function &fgr;(k)for the interference between a reference mirror plane and the film deposited patterned structure. Compensation for the phase change effect was achieved by measuring the 3 dimensional film thickness information separately prior to surface profile measurement. Then the final thin film surface profile information was measured by using the total phase function obtained through spectral frequency domain signal processing. This total phase calculation algorithm is based on spectral carrier frequency concept. © 2003 American Institute of Physics

ISSN:0094-243X    

DOI:10.1063/1.1622496

出版商:AIP    

年代:1903

数据来源: AIP

摘要:

Semiconductor microlithography is rapidly reaching a point where it becomes exceedingly difficult to shrink features at historical rates. We will no longer be able to increase process windows by going to shorter wavelengths with optical lithography, because we are running out of useable wavelengths. This necessitates either the implementation of processes with very small process windows or a transition to radically new types of lithographic technologies. Either situation presents numerous challenges to lithographers and metrologists. Particularly daunting are the requirements for gate linewidth control for microprocessors. Reducing variation requires improvement in the components of variation, each of which must be smaller than the total result. In order to improve a particular parameter, such as CD variation, metrology must be adequate for identifying improvements in the components of that parameter, not just the total. This places very tight requirements on metrology capability. Departing from optical lithography into the Brave New World of Next Generation Lithography will necessitate new metrology capabilities in several areas, not just the measurement of features on wafers. Creating the capabilities that will be needed in the future requires that funding be available for the requisite development. The need for huge amounts of funding to develop new lithographic technologies will likely necessitate a slowing down in the pace at which we shrink features. It is absolutely essential that a balance is re‐established between the prices that purchasers of chips are willing to pay and chip development and manufacturing costs. This will be very challenging with 300 mm wafer fabs coming on‐line, since low chip prices have historically been associated with overcapacity in the semiconductor industry, and it is anticipated that new lithographic technologies will be very expensive. © 2003 American Institute of Physics

ISSN:0094-243X    

DOI:10.1063/1.1622497

出版商:AIP    

年代:1903

数据来源: AIP

摘要:

Extreme Ultraviolet (EUV) lithography is under development to succeed 157nm lithography for commercial IC manufacturing for the 45nm technology node as defined by the ITRS 2001 Lithography Roadmap. EUV masks pose many manufacturing and technical challenges to meet the future commercial needs. Although some EUV mask manufacturing processes can be extended from current optical masks, many new issues arise due to transitioning to all reflective multi‐layer mirror system with patterned features versus conventional optical masks. EUV lithography operation at 13.4 – 13.5 nm wavelength requires optimal multi‐layer performance including peak reflectance, wavelength matching to the optical system, and very low defect levels. The Low Thermal Expansion Material (LTEM) that is used as a substrate for the multi‐layer reflector also requires demanding performance levels including Coefficient of Thermal Expansion (CTE), flatness, and roughness to support EUV mask needs. Performance improvements as large as several orders of magnitude are needed for some of these parameters. To aid these developments, specialized metrology tools are needed. These tools fall into two categories: Manufacturing process inspections tools include flatness interferometry, atomic force microscopy, EUV reflectometry, defect inspection, and others. Analytical tools include scanning electron microscopy, X‐Ray Diffraction, ion beam milling, and others used for problem solving. Both metrology types will play a major role in the successful development of EUV masks to meet the 45nm node requirements. This paper will review many applicable metrology techniques in addition to those listed, describe the application to EUV mask blank development or manufacturing, show problem solving examples of the techniques, and highlight particular problems or areas of need. © 2003 American Institute of Physics

ISSN:0094-243X    

DOI:10.1063/1.1622498

出版商:AIP    

年代:1903

数据来源: AIP

摘要:

Scanning Electronic Microscopes (SEM) are widely used either for cross‐section measurement (process development) or for top‐down CD measurement (production). ITRS roadmap for CD and overlay metrology points out some difficult challenges for next technology nodes. Up to now, we remained confident respectively in SEM capability and in bright field microscopy for next node requirements. Today, the limit of 0.1 micron is crossed and new requirements are stated for new 300 mm fabs. Demands arise for tighter precision, complex profile metrology and fully non‐destructive control for in‐line and integrated tools. Scatterometry is an alternative solution for CD and overlay metrology, recently introduced in fabs. We propose to review the capability of scatterometry actually demonstrated in fab and potential extensions targeting 65nm technology node. The use of optical CD tools in a production environment has also been assessed for various applications. Finally, an overview of capability extension will be given. © 2003 American Institute of Physics

ISSN:0094-243X    

DOI:10.1063/1.1622499

出版商:AIP    

年代:1903

数据来源: AIP

摘要:

Lithography is one of the most important semiconductor micro‐fabrication technologies that form mask pattern images onto the substrate. Since a mask is the original edition of semiconductor patterns, precise control of the mask aperture size becomes critical. The masks have to be made up in the accurately controlled patterns and zero defects. Therefore, mask inspection and metrology that guarantee the mask qualities are important key technologies for realizing the semiconductor production with high reliability and high yield. The advanced inspection and metrology are being developed. The requirements, technical issues, and current status of these technologies are reported. Mask inspection technologies for next generation lithography such as electron projection lithography (EPL) and extreme ultraviolet lithography (EUVL) are also reported. © 2003 American Institute of Physics

ISSN:0094-243X    

DOI:10.1063/1.1622500

出版商:AIP    

年代:1903

数据来源: AIP