Optics: measuring and testing – By polarized light examination – Of surface reflection
Reexamination Certificate
2003-01-13
2003-12-16
Sikder, Mohammad (Department: 2872)
Optics: measuring and testing
By polarized light examination
Of surface reflection
C356S370000, C356S370000, C356S370000
Reexamination Certificate
active
06665071
ABSTRACT:
TECHNICAL FIELD
The subject invention relates to a method for evaluating the characteristics of shallow junctions formed on semiconductor wafers. More specifically, the invention relates to the optical inspection of shallow junctions and the determination of energy and dose of the implants used to create the shallow junction.
BACKGROUND OF THE INVENTION
The use of optical inspection methods to evaluate ion implants has been well known for some time. Successful measurements have been made with equipment in which an intensity modulated pump beam is used to periodically excite a small spot on the sample surface. The effects of the periodic excitation, which tend to generate thermal and/or plasma waves, are monitored with an optical probe beam. One such effect is periodic surface displacements which can be measured through interferometry or by monitoring periodic angular displacements of the probe beam. Another effect is periodic changes to the reflectivity of the sample which are monitored by measuring periodic changes in the power of a reflected probe beam. Further details of such systems can be found in U.S. Pat. Nos. 4,522,510; 4,636,088; and 4,854,710.
These systems were capable of adequately measuring a relatively wide range of ion implant dosage levels. In order to improve sensitivity to higher doses, various other approaches have been taken. In one approach, the steady state reflectivity of one or more single wavelength probe beams was measured and combined with the thermal wave data to reduce ambiguities. Such an approach is described in U.S. Pat. No. 5,074,669.
Additional efforts to increase the measurement capabilities of such systems included varying the distance between the pump and probe beam spots; varying the modulation frequency of the pump source; and combining the thermal wave data with other measured data such as from spectroscopic reflectometry or ellipsometry. Such efforts are described in U.S. Pat. No. 5,978,074 and copending U.S. patent application Ser. No. 09/499,974, filed Feb. 8, 2000. All of the above cited patents and patent applications are incorporated by reference.
The above described techniques do not function to measure ion concentration directly, rather, they measure the damage done to the crystal lattice structure by the implanted ions. Variations in dosage level produce different levels of damage which can be detected by the thermal wave measurements. Variations in the energy used to implant the ions also affects the extent of damage to the lattice. As the energy level is increased, the ions are driven deeper into the lattice and the damage is more extensive.
It would be desirable to develop a measurement method which could separate out the contributions of the dose and energy levels of the implants to the damage of the wafer. In this way, the process used to create the implants can be better controlled. Such a measurement would extremely useful in the fabrication of shallow junctions in semiconductors.
More specifically, in the effort to achieve further miniaturization of semiconductor devices, the junctions dimensions must be reduced, both in width and depth. According to the 1999 SIA international roadmap, the next technology node to be achieved in two years is characterized by a lateral channel length of 130 nm, which means that the vertical drain and source pn-junction depths have to be shallower than 100 nm. Low energy ion implantation (<5 keV) has been developed to achieve these ultra-shallow junction depths,
The need to create these shallow junctions requires unprecedented control of the ion implantation process. Any unexpected variations in either dosage level or energy of the implant can result in the failure of the circuit. Therefore, it would be highly desirable to adapt the prior measurement approaches to evaluate both dosage level (ion concentration) and the energy of the implants.
Research experiments have concentrated on using destructive methods such as secondary ion mass spectrometry (SIMS) transmission electron microscopy (TEM) and spreading resistance depth profiling. Some attempts for non-destructive analysis have been made with ion scattering and spectroscopic ellipsometry, while the non-destructive thermal wave methods have demonstrated low sensitivity for implants below 5 keV.
Most SIMS equipment have a physical limitation for accurate depth profiling of ultra-shallow junctions. A transient region down to 100 Å depth is typically formed at the oxygen bombarded surface due to ionization effects at the oxidized silicon surface. Special test samples are typically required with a silicon capping layer to avoid the surface effect. TEM imaging involves tedious cross-sectional sample preparation, but is generally considered the most accurate way to measure the crystalline damage depth. Spreading resistance depth profiling requires an electrical contact to be established to the wafer surface. Specialized probe conditioning and sample preparation are needed for reliable measurement of ultra-shallow junctions and currently only a few labs have succeeded in these analyses. The ion scattering methods are restricted to give the depth distribution of the displaced silicon atoms only and have been found to lack the sensitivity to detect defects at levels which are important in device operation. Spectroscopic ellipsometry has been used with simple 1-2 layer models with effective medium approach for layer mixing, which complicates the analyses as separate recipes are needed for high and low (<2.5 keV) ion implants.
SUMMARY OF THE INVENTION
In accordance with the subject invention, these problems are overcome by combining the outputs from both a thermal wave type measurement and a spectroscopic measurement, either broadband spectroscopy (reflectometry) or broadband ellipsometry. This approach can provide accurate depth profiling of both the crystalline damage and the implanted ion distribution right after the implantation, before annealing.
In the preferred embodiment, the sample is periodically excited using an intensity modulated pump beam. A separate probe beam monitors the effects of the periodic excitation. As noted above, this can include effects such as the a) modulated optical reflectivity (MOR); b) vertical displacements (interferometry); c) angular displacements of a displaced probe beam; and d) periodic ellipsometric effects. (See, for example, PCT WO 00/68656, incorporated herein by reference.) The selected measurements produce first output signals that are supplied to a processor.
In accordance with the subject invention, a second, spectroscopic measurement is made within the same region of the sample. This spectroscopic measurement can be a reflectometry measurement at multiple wavelengths or a spectroscopic ellipsometry measurement at multiple wavelengths. The selected measurements produce second output signals that are supplied to a processor.
The first and second signals are combined in the processor to evaluate both the dosage level and the energy of the implant. More specifically, a theoretical model is set up which corresponds to the actual sample, including a substrate and the shallow junction. The model includes various characteristics of the material, for example, thickness of the damaged region, index of refraction, and extinction coefficient. The model is typically seeded with initial parameters of the sample. Using the Fresnel equations, calculations are performed to determine expected measurement data if the modeled sample actually existed and was measured. This calculated data is then compared to the actual measured data. Differences between the calculated data and the actual measured data are then used to vary the expected characteristics of the sample of the model in an iterative regression process for determining the actual composition of the sample, including the dosage level and energy of the implant.
Depending upon the particular sample, additional measurements may also be made to improve the analysis. For example, it may be desirable to use both a reflectometry measurement and a spectroscopi
Hovinen Minna
Opsal Jon
Sikder Mohammad
Stallman & Pollock LLP
Therma-Wave, Inc.
LandOfFree
Method for determining ion concentration and energy of... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method for determining ion concentration and energy of..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method for determining ion concentration and energy of... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3157333