Optics: measuring and testing – By light interference – For dimensional measurement
Reexamination Certificate
2001-07-31
2003-04-22
Allen, Stephone (Department: 2878)
Optics: measuring and testing
By light interference
For dimensional measurement
C356S502000, C073S655000
Reexamination Certificate
active
06552799
ABSTRACT:
TECHNICAL FIELD
The present invention enables the on-line measurement of semiconductor wafers or integrated circuits with high lateral resolution. In particular, the invention can be used to estimate the thickness profile of a sample consisting of a multilayer thin film on a substrate. The invention can also be used to estimate thermal properties of a sample, or quantities that influence these properties such as the ion implantation dose in semiconductor wafers. Moreover, the present invention can determine properties of a sample non-destructively and without contact with the sample, and can be used in on-line conditions on product wafers.
There are also many other possible industrial applications of the present invention. Whenever thin films or layers are deposited on a sample, or whenever properties of a sample need to be measured on short length scales, this invention can provide a wide range of information about the physical properties of the sample, such as mechanical or thermal properties crucial to the performance of the final product.
BACKGROUND ART
The non-destructive measurement of physical properties on micron, sub-micron, or nanometer length scales with optical techniques has been proposed in a wide variety of contexts.
For example a method for measuring the thickness or sound velocity of micron or sub-micron thin films has been proposed by J. Tauc et al. (see U.S. Pat. No. 4,710,030). This relies on the excitation of short wavelength stress pulses in a sample with ultrashort duration pump optical pulses and the detection of stress-induced changes in the optical constants of the sample using delayed probe optical pulses, in particular by measuring the changes in intensity of the probe beam reflected from the sample.
However, this method fails when the optical constants of the sample do not vary significantly with stress or strain in the sample at the wavelength of light used.
A similar method, that can be used to get round this limitation, was proposed by O. B. Wright and K. Kawashima (see Phys. Rev. Lett. vol. 69, 1668-1671 (1992)). It is based on the angular deflection of a probe beam arising from surface motion. Stress or strain pulses are always accompanied by motion of the sample surface or sample interfaces, and so this method can also be used to measure the thickness or sound velocity of thin films. This method, simultaneously sensitive to both phase and intensity changes in the light reflected from the sample, requires that the ultrashort optical pump and probe beams should be focused to slightly different positions at the sample surface. However, this method has the disadvantage of being very sensitive to the alignment of the pump and probe beams, in particular to the pump and probe optical spot separation on the sample. Moreover, the resolution is limited by the pointing stability of the laser used and by mechanical vibrations.
Another related method to generate and detect stress pulses with ultrashort optical pulses, proposed by T. F. Crimmins et al. (see Appl. Phys. Lett. vol. 74, 1344-1346, 1999) and H. J. Maris et al. (U.S. Pat. No. 5,864,393), relies on a grating technique making use of two pump beams at oblique incidence. This method, simultaneously sensitive to both phase and intensity changes in the light reflected from the sample, requires that two pump beams are simultaneously focused to the same spot on the sample at non-normal incidence.
However, measurements at normal incidence are highly desirable because the spots are circular and also because a single microscope objective can then be used to focus all the optical beams onto the sample, facilitating alignment and the obtention of a small optical spot size. Moreover, this grating technique requires several optical fringes to be produced on the sample at the same time, and this constitutes another reason why the method is not suitable for the highest lateral resolution measurements with the smallest optical spot sizes.
Another example of the use of optical techniques for the measurement of physical properties on short length scales is the use of thermal waves in the measurement of film thickness or ion implantation dose. A. Rosencwaig et al. (see Appl. Phys. Lett., vol. 46, 1013-1015, 1985), W. L. Smith et al. (see Nucl. Instrum. and Methods Phys. Res. Sect B, vol. B21, 537-541, 1987) and J. Opsal et al. (see U.S. Pat. No. 5,074,669) propose using chopped CW light beams to periodically excite thermal waves in a sample and to probe the resulting change in optical constants with a probe beam. This method can be used to probe to depths into the sample to within the thermal diffusion length, and is typically used for probing up to chopping frequencies in the 10 MHz region. However, at higher frequencies above 1 GHz, the method becomes impractical because the periodic temperature changes, dependent on the energy deposited in the sample by the excitation beam in one chopping cycle, decrease rapidly with increasing frequency. Higher frequency measurements are useful, because the thermal diffusion length, proportional to 1/{square root over ( )}f (f the chopping frequency), becomes smaller and the signals then become more sensitive to thermal properties of the sample in the near-surface region. When films of nanometer order in thickness are deposited on a substrate, it is, however, desirable to use such higher frequencies or equivalent short time scale pulsed optical measurements. Such short time scale pulsed optical measurements were proposed by C. A. Paddock and G. Eesley (see Opt. Lett., vol. 11, 273-275, 1986) using ultrashort pump and probe optical pulses. By monitoring the changes in optical intensity of a probe pulse reflected from the sample, it was possible to obtain a measure of the temperature change of the sample in the near-surface region in response to a pump pulse, and from the characteristic decay time of this temperature change estimate the thermal diffusivity of the sample. However, this method relies on the coupling of the optical reflectivity of the sample to temperature changes. If the optical reflectivity of the probe beam does not depend significantly on the temperature changes of the sample, the method fails.
Many optical interferometers have been proposed to measure the ultrafast changes in the properties of a sample in response to excitation by optical pulses on picosecond time scales. Such measurements have the advantage of being simultaneously sensitive to both phase and intensity changes in the light reflected from the sample, allowing a maximum of information to be extracted. If the optical reflectivity does not depend significantly on temperature or stress, there remains the possibility of monitoring temperature changes or the presence of stress pulses from changes in the optical phase. In addition, interferometric techniques are very sensitive, and interferometers can be easily calibrated to give quantitative information about, for example, the amplitude of the motion of the sample surface.
One example of an interferometric technique is time-division interferometry, in which an ultrashort optical probe pulse and an ultrashort optical reference pulse separately interact with a sample. For example, if the reference pulse is set to always arrive before a corresponding pump pulse and the probe pulse is set to always arrive after a corresponding pump pulse, the effect of the pump pulse on the sample can be determined by interfering the probe and reference pulses at a detector. Such a configuration, based on a Mach-Zehnder interferometer, was proposed by L. Sarger et al. (see J. Opt. Soc. Am. B, vol. 11, 995-999, 1994). This interferometer was designed for measurements in transmission through a sample. The Mach-Zehnder configuration is not a common path configuration, because it requires the probe and reference beams to travel completely different optical paths. It is thus is susceptible to unwanted noise sources such as mechanical vibrations or temperature changes.
Another configuration based on a Sagnac interferometer was proposed by M. C. Gabriel et al. (see Opt. Lett.,
Hurley David H.
Matsuda Osamu
Wright Oliver B.
Allen Stephone
Hill Bradford
Japan Science and Technology Corporation
Lorusso, Loud & Kelly
LandOfFree
Two-arm sagnac interferometer with two beam splitters does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Two-arm sagnac interferometer with two beam splitters, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Two-arm sagnac interferometer with two beam splitters will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3007143