Gas separation: processes – Chromatography
Reexamination Certificate
1999-03-17
2001-04-03
Spitzer, Robert H. (Department: 1724)
Gas separation: processes
Chromatography
C073S023220, C073S023360, C096S103000
Reexamination Certificate
active
06210465
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates to chromatographic analysis systems and, more particularly, to a method and apparatus for identifying components within a known sample through pattern recognition between an array of recognition coefficients representing attributes and stable relationships of historical chromatographic data and similar recognition coefficients representing attributes and stable relationships of a plurality of peaks within a chromatogram of a sample being evaluated.
BACKGROUND OF THE INVENTION
Chromatography is a known method of analyzing a sample comprised of several components to qualitatively determine the identity of the sample components as well as quantitatively determine the concentration of the components.
A typical gas chromatographic apparatus includes an injection port into which the sample is injected and mixed with an inert gas at high temperature, a column through which the various dissolved components of the sample will travel at a rate related to the characteristics of the specific components, and a detector for observing the elution of each component. The time between the injection of a sample and the observed maximum of a peak representing a specific component is called the retention time for that component. The results of a chromatographic separation are displayed as a plot of detector signal versus time, commonly known in the art as a chromatogram. A chromatogram typically comprises a plurality of peaks wherein each peak corresponds to a certain component of the analyzed sample. The area of the peak is to some degree characteristic of the amount of the respective component present in the sample. In order to insure a reliable qualitative and quantitative analysis of the sample, it is necessary that software running on, or in conjunction with the chromatograph, perform proper identification of those peaks in the chromatogram that represent certain compounds present in the sample.
As various components can have different retention times, the chromatogram will usually provide a series of sample peaks wherein each peak represents a respective component in the sample. Ideally, a chromatogram of a sample containing a plurality of components should have a respective plurality of clearly separate and identifiable peaks. Historically, chromatograms have been analyzed to determine the identity of the respective components by noting the time occurrence of the maximum value of each sample peak (retention time) and comparing the observed retention time of a sample peak in the chromatogram to a characteristic retention time for a known, or standard, peak that is derived from a standard mixture of known components. The observed retention time at which each sample peak occurs may be compared to the characteristic retention times of compounds in a standard mixture so as to allow one to assign, or name, an identity to the components in the sample under investigation.
Some conventional chromatographic equipment incorporates peak identification algorithms for improving the process of assigning compound names to the observed chromatographic peaks as a function of the observed retention times. These algorithms are based upon identification of a given peak within a computed retention time window based on one or more adjustable parameters. In particular, if the observed retention time of a given peak in a sample falls within a window centered on a characteristic retention time associated with a compound, the respective compound name is assigned to that peak. If a plurality of peaks occur within the window, the peak nearest the known retention time or the largest peak within the window for the compound is assigned the compound name.
In U.S. Pat. No. 5,905,192, and hereby incorporated by reference, a series of retention time windows are created in accordance with standard peaks. Each retention time window is determined according to the location of its respective standard peak and according to steps for determining one or more shift, stretch, and distort parameters. One or more of the shift, stretch, and distort parameters are used to position each retention time window. The retention time windows are applied in a fashion such that each standard peak is compared to each of the sample peak(s) captured in a given window so as to determine one or more matched pairs of peaks. That is, for each peak in the standard peak group that corresponds to a certain one of the peaks in the sample peak series, a matched pair is determined. A quality factor representing the correspondence of the standard peak to each of the candidate sample peaks in the matched pairs is determined. The resulting quality factors may be compared to determine a best match of the standard peak to one of the candidate sample peaks. The candidate sample peak in the best matched pair is then identified with the compound associated with the respective standard peak. Area percent is also mentioned as a comparison method.
Several shortcomings remain in the above-described approaches. Firstly, they all rely primarily on time domain information, in particular, retention time, for peak identification to succeed. Furthermore, adjustable parameters must be set by those with “some” skill in the practice of chromatography. To increase repeatability over an extended series of separations relatively expensive chromatographic equipment and experienced operators are required to analyze a chromatogram, adjust integration and analysis parameters (including identity parameters) so as to obtain proper identification of sample components.
SUMMARY OF THE INVENTION
Every chromatographic peak has attributes that describe its size, shape and location. While retention time is a location attribute (used extensively in the prior art), the invention provides for the utilization of additional attributes and stable relationships to aid in the identification of unknown components. For example, area and height of a peak are attributes of size; width and symmetry describe a peak's shape, as do algorithmic measures of peak type. It has been found that the attributes of a desired peak (including retention time) are not strictly independent, and may interact. The invention uses functions (which may be non-linear) of several variables to describe how the attributes (by which a desired peak can be recognized) may vary over an analytical range of interest. Coefficients indicating the relationship of a particular peak with its neighbors are also of utility for identification purposes. A simple example is the relative retention time between peaks (which can be expressed as the coefficient C
IJ
=RT
I
/RT
J
). The use of such relationship information renders some immunity to the expected experimental problems associated with chromatography that are not addressed by a dependence on retention time alone.
According to the teachings of the present invention, a method is provided for identifying components within a known type of sample through pattern recognition between an array of recognition coefficients generated from the attributes and stable relationships of a plurality of peaks within a chromatogram of a sample to be analyzed and a similar array of recognition coefficients generated from the attributes and stable relationships of a plurality of peaks from a set of one or more historical chromatograms of the known type of sample. The group of attributes include, but are not limited to: retention time, peak shape including symmetry, width and duration, peak size including area and height, and algorithmic measures of peak type. Peak type combines shape and size attributes indicating both the gross appearance of a peak (for example, providing for classification as a solvent peak or as a tangent peak), and fine details such as the starting and ending baseline assignments
When identifying components within a sample, the invention relies on the presence of certain required peaks in the sample chromatogram to ascertain the general relationships between desired peaks representing the components of the sample to be identified.
Agilent Technologie,s Inc.
Spitzer Robert H.
LandOfFree
Method for identification of components within a known sample 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 identification of components within a known sample, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method for identification of components within a known sample will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2533889