Optics: measuring and testing – By dispersed light spectroscopy
Reexamination Certificate
2001-08-02
2004-01-06
Evans, F. L. (Department: 2877)
Optics: measuring and testing
By dispersed light spectroscopy
C356S326000
Reexamination Certificate
active
06674526
ABSTRACT:
FEDERALLY SPONSORED RESEARCH
Not applicable
SEQUENCE LISTING OR PROGRAM
Not applicable
SEQUENCE LISTING
Not applicable
REFERENCES CITED
1. R Marbach, On Wiener Filtering and the Physics Behind Statistical Modelling, Journal of Biomedical Optics 7(1), pp. 130-147 (January 2002).
FIELD OF THE INVENTION
The invention relates to methods and apparatus for improving the long-term stability of spectroscopic quantitative analyses.
BACKGROUND OF THE INVENTION
Spectroscopic quantitative analysis (SQA) is the spectrally resolved measurement of light interacting with a sample for the purpose of quantifying a property of that sample. A simple example is an infrared (IR) transmission measurement through a cuvette filled with a liquid sample, for the purpose of quantifying a component concentration in that sample.
A multitude of different optical phenomena, e.g., absorbance, scatter, fluorescence, rotation of polarization, etc.; exist and can be used for SQA, and the nature of the sample and its property of interest will usually determine which phenomenon is best to use for a particular measurement application. Many applications work in the visible wavelength range, i.e., 400-700 nm (VIS), but the majority of applications require the measurement of light at wavelengths outside of the VIS, and typical wavelength ranges employed range from the deep ultraviolet (&lgr;<200 nm) to the long-wave infrared (&lgr;>20,000 nm). A multitude of different hardware configurations exist for resolving the optical spectrum, e.g., optical filters, interferometers, optical gratings, matched-wavelength LED's, etc.; and selection depends on the particular application, the wavelength band, the required accuracy of the measurement, as well as various marketing factors. This invention applies to all optical phenomena and to all hardware configurations, from the simplest case of a single optically resolved wavelength “band” realized by the shape of the emission spectrum of the selected light source, to the most complicated and expensive configurations resolving thousands of wavelength bands.
Typical application areas for SQA are industrial process control (e.g., chemical process optimization) and industrial quality control (e.g., incoming material checking) and in recent times also medical applications (e.g. blood analysis) and consumer applications (e.g. food freshness control or indoor gas sensing). A common characteristic of all SQA measurements is the fact that the measurement is actually a two-step process. First, an optical spectrum is measured and second, a calibrated algorithm is applied to the measured data to transform the spectrum into the desired value of the property of interest. The second step is typically performed in software and is therefore not subject to long-term drifts or temperature instabilities. The long-term reliability of the whole system is almost always limited by the instability of the instrument hardware.
Long-term drifts and temperature instabilities of the hardware affect both the x- and the y-dimension of the spectra. Without any loss of generality, assume that optical wavelength (x-dimension) is measured in, e.g., nanometers (nm) and that the amplitude of the detected spectral signal (y-dimension) is measured in, e.g., Volts per nanometer [V
m].
Typical causes of amplitude instability include lamp-aging, accumulation of dirt on the optics, electronic drifts in the amplifiers, and responsivity fluctuations of the photodetector. When propagated through the computer algorithm, amplitude instabilities equate to measurement errors in the property of interest, with the proportionality factor being strongly dependent on the spectral shape of the instability effect.
A typical cause of wavelength axis instability is thermally induced strain in the optomechanics of a spectrograph. Wavelength axis instabilities transform themselves into equivalent amplitude instabilities first, and then into measurement errors. The fundamental difference between amplitude instabilities and wavelength axis instabilities is that amplitude instabilities are ratioed out by spectroscopic referencing, see the discussion below.
The computer algorithms typically need processed ratios between a sample spectrum and a reference spectrum as input. In the following text of the specification and the claims, these input spectra will generally be referred to as “absorbance spectra” regardless of the underlying optical measurement principle, e.g., transmission or diffuse reflection.
In a general sense it can be said that amplitude instabilities are typically the dominant problem in simple SQA applications, i.e., in well-posed measurements with relatively large spectral signal-to-noise ratios (SNR). “Well-posed” here means that the root-mean-square (RMS) variation of the spectral fingerprint of the signal is comparable to, or even larger than, the RMS variation of even the largest eigenfactors of the spectral noise (where “largest eigenfactor” is short for “eigenfactor with the largest eigenvalue”)[1]. The technique of spectroscopic referencing is usually performed to combat amplitude instabilities, e.g., in the case of the transmission cuvette, the spectrum from the sample S(&lgr;)[V
m] can be ratioed by a reference spectrum R(&lgr;)[V
m] from, e.g., the empty cuvette or no cuvette at all (“empty path”), to form an absorbance spectrum A(&lgr;)=−log
10
(S(&lgr;)/R(&lgr;))(in units of [AU]). In some applications, e.g., in a diffuse reflection-type measurement, it is not possible to use the empty cell for the reference measurement, and a dedicated reference element must be placed into the path of the measurement light, typically at the same position where the sample is located during the sample measurement. Design, choice of material, and handling of the reference element are all usually optimized in order to maximize the long-term stability of the optical response of the reference.
Referencing is almost universally employed because it is well suited to decrease amplitude instabilities in the larger eigenfactors of the spectral noise, and is often sufficient to achieve satisfactory long-term performance in simple SQA applications.
Many of the more recent SQA applications, however, are precision measurements, i.e., ill-posed measurements where the RMS variation of the spectral fingerprint of the signal is smaller than the RMS variation of many of the larger eigenfactors of the spectral noise. Graphically speaking, the spectral signal can no longer be seen with the naked eye when the spectra are plotted. In these applications, the required spectral SNR does not reside in the spectral space spanned by the larger eigenfactors, but has to come from the smaller eigenfactors [
1
]. Wavelength axis instabilities are a significant contributor to the noise in the smaller eigenfactors, and are especially detrimental to the measurement because they can not be referenced out by using the spectroscopic referencing techniques of the prior art.
SUMMARY OF THE INVENTION
This invention provides methods and apparatus for improving the long-term stability of spectroscopic quantitative analyses, by eliminating or significantly reducing the detrimental effects of wavelength axis instabilities. The effects of wavelength axis instabilities are diminished by inserting an optical element into the path of the measurement light, which is called the inverse sample element (ISE) and which has spectral characteristics designed such that wavelength axis instabilities of the instrument hardware cause opposite and nearly cancelled amplitude effects from the sample and the ISE in the resulting absorbance spectrum.
REFERENCES:
patent: 5422719 (1995-06-01), Goldstein
patent: 5784158 (1998-07-01), Stanco et al.
patent: 5822071 (1998-10-01), Dosmann et al.
patent: 6404970 (2002-06-01), Gransden et al.
Evans F. L.
Geisel Kara
LandOfFree
Methods and apparatus for improving the long-term stability... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Methods and apparatus for improving the long-term stability..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Methods and apparatus for improving the long-term stability... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3247909