TECHNICAL FIELD

[0001] The present disclosure relates generally to blank runs and system performance in the field of chromatography.

BACKGROUND

[0002] Chromatography is a method of analyzing a sample comprised of one or more components to qualitatively determine the identity of the sample component(s) as well as quantitatively determine the amount of the component(s). While the ensuing discussion focuses on gas chromatography, the concepts are extensible to all separation techniques.

[0003] A typical gas chromatograph (GC) includes an injection port or inlet into which a sample is injected, a column through which the various components of the sample will travel at a rate related to the characteristics of the specific components, an oven or other type of heating and/or cooling device to control the temperature of the column, and a detector for observing the elution of each component. A carrier gas, or mobile phase, carries the sample from the inlet, through the GC column, and to the detector. The column contains a stationary phase which separates the sample components based on their differential affinity for this coating. GC columns can be packed with stationary phase containing particles (packed columns) or can be made with small bore open tubes coated with stationary phase (capillary columns). Liquid samples can be injected into the inlet via an autosampler or manual syringe. Other examples of sample introduction may include headspace sampling, thermal desorption, gas or liquid sampling valves, or solid sample introduction devices.

[0004] What is known as a GC method specifies the parameters associated with the operation of the GC for a particular analysis. Some of the parameters include flow rates, pressures, temperatures of the various GC components, injection parameters, as well as others. A sequence refers to a set of analyses to be run serially, each analysis being run according to a GC method. The results of a chromatographic separation are displayed as a plot of detector signal versus time, commonly known in the art as a chromatogram. When a sample component or components elute from the column into the detector, the detector signal will change from its nominal value. When the component or components have finished eluting from the column and passed through the detector, the detector signal level will return to its original value or very close to it. This resulting deflection is typically shaped as a sharp bell-shaped curve and is referred to as a peak. A chromatogram typically comprises one or more peaks, each corresponding to a certain component of the analyzed sample. 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 area or height of the peak is to some degree characteristic of the amount of the respective component present in the sample and can be calculated by integrating the peak.

[0005] The chromatogram contains a wealth of information in addition to the peak retention time and area. The continual performance of the chromatographic system can be assessed by reviewing certain parameters of the chromatogram. The value of this assessment has been known for many years and is typically referred to as system suitability, system validation, or system performance. This concept was formalized into pharmaceutical development in the l980s-l990s, but was also well known in other separation science applications. As a reference, the FDA Center for Drug Evaluation and Research (CDER) published a monograph entitled“Reviewer Guidance - Validation of Chromatographic Methods” in November of 1994. In this document, the system suitability specifications and tests are described in section IV, subsection J.

[0006] Normally, the sample used to determine system suitability contains the component or components of interest at an easily detectable amount and any additional interferences that may be observed in the real sample. For example, in a pharmaceutical formulation analysis, the sample may contain additional components as well as the active drug ingredient. These additional components may elute close to the drug component and the resolution of these two components needs to be monitored. Subsection J lists the following criteria for assessing the performance of the GC method:

Capacity factor (relative retention time of the compound with respect to the system void time);

Precision/Injection repeatability (consistency of the chromatographic system to produce a valid result);

Relative retention (relative retention of the compound of interest compared to the retention of another compound in the sample);

Resolution (measurement of how well two peaks are separated);

Tailing Factor (measurement of distortion of the compound peak); and

Theoretical plate number (measurement of how efficient the chromatographic system is operating with respect to compound transport).

[0007] Since that initial monograph, other parameters have also been used to evaluate the performance of the chromatographic system, such as: Peak width (indication of efficiency and column/system degradation);

Peak area (indication of detector sensitivity);

Peak response factor (relative sensitivity compared to a standard component in the sample);

Skew (third order peak moment analysis); and

Kurtosis (fourth order peak moment analysis).

[0008] One paradigm for evaluating system suitability is to run a reference or quality control sample and evaluate the resulting chromatogram. Some or all of the parameters listed above, or even additional parameters not listed in the FDA monograph, are determined and serve as reference values for subsequent analyses. Multiple replicate runs are made to establish the statistical precision of the measurements and determine the pass/fail criteria for determining if the system is performing properly. Acceptance criteria are established by the user for a particular analysis. As a specific example, the retention time of a particular compound is determined to be 5.00 +/- 0.01 minutes. In subsequent system suitability analyses, if the measured retention time does not fall into the acceptable range, the chromatographic system is deemed out of compliance and maintenance may be required to restore performance. Acceptance criteria can be either one sided (performance parameter must be greater than or less than a user determined or regulated chromatography method value) or two sided as typically seen in control charting. For example, suggested criteria are provided in the FDA monograph, such as repeatability of peak area or peak response should be better than 1% RSD for five replicate analyses.

[0009] This concept of using a reference sample to assess instrument performance has also been used in a number of prescriptive standardized chromatographic methods (e.g. US EPA 8270, GC/MS analysis of semi-volatile organic compounds). This reference sample is analyzed during a sequence of analyses to verify that the instrument is still complying with its performance metrics. If the instrument fails this performance check, the user is required to stop the sequence and repair the instrument to bring it back into compliance. As an example, for US EPA method 8270C (published in 1996), the calibration check sample must meet criteria for:

Response factor: between 80% and 120% of original value

Internal standard response: between -50% and +100%

Internal standard retention time: +/- 30 seconds

Relative retention time within an acceptable range (+/- 0.06)

Detector linearity response: RSD <= 15%.

[0010] From the previous description, the concept of using retention time and relative retention time (and by extension retention index) to indicate instrument performance has been well known for years.

[0011] An aspect of chromatographic analysis that has not been explored as rigorously is the use of blank runs to evaluate whether or not the system is performing as intended.

Chromatographers typically make blank runs to confirm that the system is“clean” and that components identified in the subsequent sample did in fact come from that sample and not the previous sample. However, the blank run can also be analyzed to determine the system performance or system viability as will be illustrated below. Blank runs provide different information than that obtained with a reference sample.

[0012] To assess instrument performance, there are two types of blank runs. The first type is a no-injection blank run, whereby the GC method parameters are executed, such as column flow, pressures, oven temperature, detector data collection, and others, but no injection is made. An example profile of a no-injection blank run with a single oven temperature ramp is shown in FIG. 1. Several artifacts on the baseline are evident in the no-injection blank run of FIG. 1. First, there are measurable peaks 101 showing up on the baseline that may impact quantitation of sample peaks during a sample run. Potential causes for this are chemical compounds eluting from the inlet into the column from previous runs or chemical compounds“bleeding” off the inlet liner or septum. Next, the baseline rises as the oven temperature increases. This is due to the phenomenon known as column bleed that occurs as the stationary phase increasingly breaks down as the temperature rises, leaves the column, and is sensed by the detector. This final baseline 102 can be substantially higher than the initial baseline 103 depending on the column temperature used. Lastly, the baseline noise increases at the end of the run. This may be due to the column aging and degraded stationary phase continually bleeding into the detector.

[0013] In the second case of a blank run, a“clean” solvent (one that contains no components of interest or components that interfere with the analysis) is injected while running a GC method. This is referred to as a solvent blank run. Some examples of solvents include dichloromethane, acetone, hexane, acetonitrile, and ethyl acetate. For a sample analysis, these solvents will generally elute before any components of interest and are referred to as a solvent peak 201. In addition to verifying that the chromatograph is free of interferents as evidenced by the flat baseline after the solvent elutes, this type of blank run can verify that the autosampler and inlet are also“clean” and functioning properly and that the detector response is stable. FIG. 2 shows the resulting chromatogram from a solvent blank run.

[0014] An extreme example of the use of blank runs is seen in the analysis of drugs associated with abuse. The sequence applied consists of a solvent blank run, a first real sample run, a solvent blank run, a second real sample run, followed by a solvent blank run. This gives defensible proof that the system had no carryover from the first real sample to the second real sample. An expert chromatographer provides a manual determination of whether a system was clean (that is, free from carryover, contamination, spurious baseline disturbances). This is done by manually inspecting the results of the blank run to ascertain if there are any interferences observed. However, this manual assessment is limited because, in addition to this process being time-consuming, it happens in post-processing after the samples are run. If there was an issue, the sample(s) may have to be prepared again and re-run. Additionally, the expert inspecting the data typically reviews the integration results to see if any calibrated compounds (in this case, drugs of abuse) are observed. The actual baseline is not evaluated. In some references, such as EPA method 8000D, these two types of blank runs may also be referred to as instrument blank runs. This is to avoid confusion with sample preparation blank runs (e.g. method blanks and laboratory control blanks) and other quality control samples used to validate an entire analysis protocol. For the sake of brevity, blank run is used here to mean either the no-injection blank run or a solvent blank run. The intention of a reference sample is to verify that the components of interest elute at the correct retention time, are sufficiently separated, and have the necessary detector response to comply with the analysis goals. EPA method 8000D, the EPA’s guidance document for calibration and quality control, describes these requirements in detail in section 9.3.3. The use of blank runs is described in section 4.4. While the section describes elevated baselines and the errors they can cause in quantitation, there are no criteria given to assess the data. The interpretation of blanks is given in section 9.2.6.8 to 9.2.6.12. Similarly, there is no discussion of detector signal or noise and how it might be used to evaluate the instrument viability.

[0015] Another use case is when the chromatographer wants to subtract the system contribution to baseline noise or contaminant peaks from a sample signal. As an example, the system contribution can include the detector signal increase resulting from the column stationary phase“bleeding” into the detector causing the baseline to rise. Similarly, components can be thermally extracted from the inlet liner, the column connection fittings, and the septum used to seal the inlet. These will also result in a rise in detector signal. In this case, a no-injection blank run is made first as a reference. This blank run should show just the detector baseline as a function of the system. For applications like simulated distillation, the detector signal from this blank run is then subtracted from all chromatograms of real samples to provide a flatter baseline and easier integration when determining the area of sample peaks. This can be done by utilizing the column compensation feature available on a gas chromatograph or in data post processing.

OVERVIEW

[0016] Described herein is a method for evaluating performance of a chromatographic system having a detector operable to output signals. The method includes receiving one or more blank run acceptance criteria, receiving from the detector a detector signal from a blank run, extracting attributes of the detector signal, comparing the extracted attributes against corresponding blank run acceptance criteria, and providing a decision relating to the viability of the chromatographic system based on results of the comparison.

[0017] Also described herein is a chromatographic system including a separation column, an injection port for introducing a sample into the separation column, a detector coupled to the separation column and operable to output a detector signal, and a blank run analysis unit (BRAET) operable to receive a detector signal from a blank run. The BRAET includes an attribute extractor for extracting attributes of the detector signal, and a decision module for comparing the extracted attributes against corresponding acceptance criteria and providing a decision relating to the viability of the chromatographic system based on results of the comparison.

[0018] Also described herein is a blank run analysis unit (BRAU) operable to receive from a detector of a chromatographic system a detector signal from a blank run. The BRAU includes a memory for storing executable instructions, an attribute extractor, a decider; and a processor for executing the instructions to cause the attribute extractor to extract attributes of the detector signal, and to cause the decider to compare the extracted attributes against corresponding acceptance criteria and provide a decision relating to the viability of the chromatographic system based on results of the comparison.

[0019] Advantages of certain embodiments disclosed herein include freeing the user from having to open and visually inspect all of the blank run data to confirm the data accuracy. This tedious process delays the completion of the chromatographic analysis and is prone to error and validity issues.

[0020] Other advantages include allowing the chromatographic system to evaluate the blank run and take a user-directed action without requiring user intervention. For example, a system with a blank run that fails the acceptance criteria for one or more attributes can pause the analytical sequence of samples to be run and wait until the issue can be rectified. By not continuing the sequence, the remaining samples will not be run with a non-compliant instrument. This saves the user from having to rerun the samples or having to prepare new samples to replace those already injected under non-compliant conditions. [0021] Other advantages include allowing the chromatographic system to pause the sequence and execute a maintenance step before resuming the sequence. Usually, a column bake-out will repair the system and permit the blank run to pass. With this capability, these tasks could be automated completely.

[0022] Other advantages include allowing the system to track the results of the blank runs and could provide control charting for the number of sample analyses before maintenance is needed.

[0023] Other advantages include allowing the system to troubleshoot automatically if a blank run analysis determines the system is not viable. For example, in a solvent blank run, the solvent peak retention time is one of the criteria to be evaluated. If the retention time is longer than expected but the column head pressure is correct, this indicates that the column is partially plugged and requires maintenance.

[0024] Further embodiments, features, and advantages of the invention, as well as the structure and operation of the various embodiments of the invention are described in detail below with reference to accompanying drawings.

[0025] As used herein,“system suitability,”“system validation,”“system performance,” “system viability” mean an assessment of the chromatographic performance of the

chromatographic system, typically using retention time, peak area, and other parameters to determine if the chromatograph is capable of performing a valid analysis.“Interferences” mean sample components that may complicate an accurate analysis of the component or components of interest.“Reference sample” or“quality control sample” means a sample containing components used to evaluate the system chromatographic performance.“Acceptance criteria” means one or more ranges or thresholds for an attribute that, if met, indicate that a

chromatographic system is viable.“No-injection blank run” means a chromatographic analysis in which no sample is introduced into the instrument but the GC method conditions are executed. “Bleed” means additional chemical elution from the column that raises the detector signal level. Bleed can come from multiple sources and is normally elevated due to higher temperatures in the system.“Clean solvent” means a chemical sample composed of just the solvent.“Solvent blank run” means a chromatographic analysis where a clean solvent or solvents are introduced into the instrument.“Extracted attributes” means the value of an attribute for a given blank run. This value can be compared to the acceptance criteria to determine if the chromatographic system is viable.“Detector signal” is the original detector output before any filtering or extraction is applied.“Baseline signal” mean the resulting signal data from applying a filter to the original detector signal data to eliminate noise and other peaks.“Residual signal” means the remaining detector signal once the baseline extraction has been completed. Typically, this will contain narrow peaks and high frequency components.

BRIEF DESCRIPTION OF THE DRAWINGS

[0026] The accompanying drawings, which are incorporated into and constitute a part of this specification, illustrate one or more examples of embodiments and, together with the description of example embodiments, serve to explain the principles and implementations of the

embodiments.

[0027] In the drawings: FIG. 1 is a chromatogram of a no-injection blank run with a single oven temperature ramp;

FIG. 2 is a chromatogram of a solvent blank run;

FIG. 3 is a schematic diagram of an example gas chromatographic system that includes a blank run analysis unit (BRAET) according to an embodiment;

FIG. 4 is a block diagram that shows a BRAET in further detail in accordance with one embodiment;

FIGS. 5a, 5b, and 5c illustrate an example of the processing of a detector signal. FIG. 5a shows the raw detector signal. FIGS. 5b and 5c illustrate the decoupled baseline and residual components, respectively. Results calculated by the attribute extractor such as initial baseline, final baseline, and a region of interest are illustrated in both FIGS. 5b and 5c;

FIG. 6a, 6b, and 6c are tables that list various attributes of interest, along with corresponding failing conditions, and interpretations for the failing conditions, in accordance with an embodiment; and

FIG. 7 is a flow diagram showing an example use case in accordance with an embodiment.

DESCRIPTION OF EXAMPLE EMBODIMENTS

[0028] Example embodiments are described herein in the context of the analysis of blank runs for evaluating chromatograph performance. Those of ordinary skill in the art will realize that the following description is illustrative only and is not intended to be in any way limiting. Other embodiments will readily suggest themselves to such skilled persons having the benefit of this disclosure. Reference will now be made in detail to implementations of the example embodiments as illustrated in the accompanying drawings. The same reference indicators will be used to the extent possible throughout the drawings and the following description to refer to the same or like items.

[0029] In the detailed description of embodiments that follows, references to“one embodiment”,“an embodiment”,“an example embodiment”, etc., indicate that the embodiment described may include a particular feature, structure, or characteristic, but every embodiment may not necessarily include the particular feature, structure, or characteristic. Moreover, such phrases are not necessarily referring to the same embodiment. Further, when a particular feature, structure, or characteristic is described in connection with an embodiment, it is submitted that it is within the knowledge of one skilled in the art to effect such feature, structure, or characteristic in connection with other embodiments whether or not explicitly described.

[0030] In the interest of clarity, not all of the routine features of the implementations described herein are shown and described. It will be appreciated that in the development of any such actual implementation, numerous implementation-specific decisions must be made in order to achieve the developer’s specific goals, such as compliance with application- and business- related constraints, and that these specific goals will vary from one implementation to another and from one developer to another. Moreover, it will be appreciated that such a development effort might be complex and time-consuming, but would nevertheless be a routine undertaking of engineering for those of ordinary skill in the art having the benefit of this disclosure.

[0031] FIG. 3 is a schematic diagram of an example gas chromatographic system 300 according to certain embodiments. An injection port 301 receives a sample (not shown) to be analyzed. The sample is carried by a carrier gas (not shown) through a separation column 302 under carefully controlled conditions of pressures, temperatures, and flows, before reaching detector 303. Column 302 could consist of one or many columns to achieve the component separation. Oven 304 is provided for heating the column to a desired temperature. Although not shown, other types of heating and/or cooling are also contemplated. Detector 303 generates measurement signals relating to various sensed parameters of the sample, and delivers these output signals to processor 305 for analysis. Although not specified, detector 303 could be any GC compatible detector. Processor 305 may be one or more microprocessors or processing units including, but not limited to, single-core or multi-core processors, or other type of logic or data processing circuitry. In certain embodiments, the analysis is carried out by the processor 305 in accordance with executable code stored in memory 306, which may be in the form of volatile (for example DRAM) or non-volatile (for example FLASH) memory, or a combination of these. A user interface (UI) 307 allows an operator to receive information from the processor, and to input information and parameters to the processor. Such information may be stored and accessed through memory 306. For example, UI 307 can include a display, keyboard, touchscreen, voice- activated input, tactile device, and/or other peripherals to provide control inputs and/or outputs.

[0032] Gas chromatographic system 300 also includes a blank run analysis unit (BRAU) 308 shown coupled to processor 305 and memory 306, and generally used to evaluate the performance or viability of the chromatographic system using blank runs. BRAU 308 may be one or more hardware, software, or firmware components, or any combination thereof, that are separate from or integral with the processor 305. In certain embodiments, BRAU 308 may include its own processor (308a) for executing instructions, and its own memory (308b) for storing instructions and data, such as data from detector 303 and from the user, including, for example, the blank run acceptance criteria further detailed below. In certain embodiments, BRAU 308 executes code that may be stored in memory 306, and generates output results that may be organized as files 309 in memory 306, discussed in more detail below. In certain embodiments, BRAU 308 may be one or more components made up of software, firmware, hardware, or any combination thereof, that is separate from the gas chromatographic system 300, for example as part of an external diagnostic device 310 (BRAU 308’) having its own processor, memory, user interface, and other supporting components (not shown).

[0033] A block diagram explaining the operation of blank run analysis unit (BRAU) 400 is shown in FIG. 4. Each block in FIG. 4 may correspond to a discrete hardware, firmware, and/or software component having the functionality described herein. Alternatively, the described functionality may be achieved by any combination of the depicted blocks, or by additional blocks or sub-blocks that are not depicted in the interest of clarity. Generally, and without limitation, BRAU 400 evaluates the performance or viability of the chromatographic system by, for example, detecting when a blank run has completed, extracting the detector signal from the blank run, evaluating the detector signal from the blank run, and determining if extracted attributes of the signal fall within acceptance criteria. In certain embodiments, BRAU 400 monitors the system and processes blank run data when it becomes available. In certain embodiments, it decouples the detector signal into a baseline and residual signal and calculates a set of values for attributes from one or both of the decoupled pair of signals and compares these extracted attributes to the acceptance criteria. In certain embodiments, it reports back the results with a PASS or FAIL, indicative of the chromatographic system’s viability and whether or not it is functioning normally.

[0034] Information relating to a blank run is provided to BRAU 400. This information includes the detector signal that is used to construct the chromatogram. Additional information, collectively shown as User Inputs in FIG. 4, that may be provided to BRAU 400 includes the GC method, region(s) of interest of the detector signal, user-defined acceptance criteria, and the type of blank run made. It will be noted that in the context of this example,“GC method” refers to the instrument parameters under which the run is made. The region(s) of interest may define the time range(s) to which to apply the evaluation of attributes where the user is concerned about specific portions(s) of the detector signal. This region of interest(s) will typically be the time range in which components of interest will elute during a sample run since this is where noise and peaks due to contamination will cause the most issues with the sample analysis. In certain

embodiments, the type of blank run can be a no-injection or a solvent blank run. Additional information can be received, including chromatographer ID, the sequence and run name or unique ID, the date and time of the run, and other instrument-specific information corresponding to the run. Operator input, via UI 307 (FIG. 3) for instance, can be used to provide some or all of this information, including, in certain embodiments, whether the blank run was a no-injection blank run or a solvent blank run. In addition, the operator input can also be made through an external device, such as diagnostic device 310 (FIG. 3), which may be capable of controlling the operation of the GC, and inputting and storing the aforementioned information.

[0035] In certain embodiments, once BRAU 400 receives and optionally validates the information, it begins processing the output signal from the detector. FIG. 5a illustrates an example of a detector signal from a no-injection blank run from a contaminated system in which a region of interest 501 from 18 to 23 minutes is specified by the user based on the analysis (in some embodiments it may be specified by the system automatically, rather than by the user). In an example implementation, the output signal from detector 303 is delivered to filter 401 and then decoupler 402, which decouples the detector signal into two separated signals containing the baseline and the residual signal. FIGS. 5b and 5c illustrate baseline and residual signals extracted from the detector signal of FIG. 5a, respectively. FIG. 5b also shows the initial baseline region 512 and final baseline region 513 that will be processed by the attribute extractor. In certain embodiments, the baseline signal is extracted first and then subtracted out of the detector signal, with the remaining signal forming the residual signal. In an ideal case, the baseline signal contains only the detector signal caused by the carrier gas eluting from the column. In reality, the baseline will show the shift in detector signal caused by anything eluting from the column over a long period of time. A prominent contributor to the baseline is stationary phase column bleed. Most GC columns utilize a high molecular weight silicone polymer that is immobilized in the column as a stationary phase. At very high temperatures, small amounts of the stationary phase can degrade and elute over an extended period of time. This creates the rise in the baseline shown at the end of the chromatographic run in FIG. 5b. The residual signal contains the deviations in detector response due to more quickly eluting material that manifest as peaks or increased, high frequency noise. Examples of the residual signal are contaminants in the carrier gas or inlet that accumulate on the inlet of the column and elute when a chromatographic run is made. The sum of these two signals yields the original detector data.

[0036] In certain embodiments, the baseline extraction is performed by using non-linear median filtering, for example at filter 401 in FIG. 4. Other types of filtering may be used as well. In the example of FIG. 5a, the detector signal was filtered with a median filter using a 40 second window to generate FIG. 5b. This filter was selected due to the relatively low computational complexity and beneficial properties. The median filter provides excellent suppression of noise from heavily tailed statistical distributions that appear as spikes or peaks in the data. This makes it a good choice to extract the baseline while ignoring sharp peaks. It also maintains baseline shifts in the data like those found in blank runs that exhibit a steep rise during the temperature ramp and then remain constant once the final temperature is reached.

[0037] Decoupling allows for the calculation of values based on attributes, or extracted attributes, involving the sum or maximum of peaks without the need for identifying each individual peak. This allows for quicker algorithm development and less computational complexity. Decoupling also permits easier peak identification, integration, and the calculation of more complex attributes. However, the approach of using a chromatographic peak integrator is a viable alternative. In this alternative approach, the original detector signal can be processed by a chromatographic peak integrator 403 to determine peak retention times and peak area or height. Peaks from the chromatographic peak integrator 403 can then be combined with those from all other peaks to determine the maximum height or total sum attribute by the attribute extractor 404 This alternative step is shown in FIG. 4 with peak integrator 403 receiving the detector signal and providing its output to attribute extractor 404 Such an approach can be used either with or without filtering of any type. In a specific implementation, a peak integrator uses the median filter to obtain a curve containing all peaks. In certain embodiments, for

computational simplicity, individual peaks are not identified and characterized separately.

[0038] It should be noted that the baseline extraction is not limited to the above described median filter and other filtering or smoothing techniques can be applied. Selection of these techniques depends on the computational complexity and speed desired by the designer, the characteristics of the noise to be suppressed, whether the signal comes from a wide-sense stationary process, and whether other a priori information is available about the detector signal (or the system) that can be exploited. Filtering techniques are generally best for applications requiring real time processing but can contain a delay or lag in their response. Other approaches such as smoothing generally have better noise suppression and can be easily implemented without response delay. However, they cannot be implemented in real time and are limited to post-processing. Other possible filters and smoothers for baseline extraction are discussed below.

[0039] Finite and Infinite Impulse Response filters and smoothers (FIR and HR) are perhaps the most well-known, simple to implement, computationally fast, and are well suited to suppress sinusoidal noise across a range of frequencies. This makes them suitable to suppress high frequency (or frequency specific) baseline noise. One particular FIR filter widely used in the field of analytical chemistry is the Savitzky-Golay filter. It is known for its ability to reduce high frequency noise from a signal while maintaining the shape and height of its peaks. However, it is not well suited to estimate the baseline signal since it is designed to preserve peaks in the signal and not filter them out. In fact FIR filters in general are not able to suppress noise that appears as peaks or noise that, more precisely, contains a wide and flat power spectral density. Ordered statistic filters, such as the median filter discussed above, are better able to suppress these peak shaped disturbances; however, they may be sub-optimal to suppress high frequency noise, periodic, or random white noise that is typically found in a chromatographic run. Adaptive filters and smoothers are useful when the characteristics of the noise are unknown or are changing over time (not stationary or wide-sense stationary). This class of filters uses mathematical or statistical models of the noise, employs optimization techniques to estimate the noise characteristics, and adapts the filter to best suppress the noise over time. The adaptive nature of these algorithms proves useful when the noise characteristics change over time; however, they are generally harder to design and require significant processing power. Any combination of smoothing or filter techniques can be used to best suppress the noise. [0040] Returning to FIG. 4, the decoupled baseline and residual signals are delivered to attribute extractor 404 for extracting attributes of each of these signals. The attribute extractor calculates values for any defined attributes for the blank run. Some of these extracted attributes may be calculated using the baseline signal, residual signal, original detector signal or from any combination of these signals.

[0041] Illustrated in FIG. 5b is an example of the attributes Initial Baseline 512, Final Baseline 513, and region of interest 511 that can be calculated by attribute extractor 404 of FIG. 4 using the baseline signal. The purpose of these calculations is to detect failures corresponding to the beginning and end of the detector signal and also the region of interest. These failures include a malfunctioning detector, if the detector is off, or if the column is broken. Initial and Final Baseline attributes, as well as other examples of attributes pertaining to the detector signal or baseline signal, are listed and defined in the table in FIG. 6a. These examples of attributes can, in whole or in part, give an indication of the viability of the chromatographic system to conduct an analysis.

[0042] Similarly, illustrations of extracted attributes of the residual signal that can be calculated by attribute extractor 404 are shown in FIG. 5c. Initial Baseline Noise 522 and Final Baseline Noise 523 calculations are made at the beginning and end of the residual signal, respectively. These can include any calculation that measures the statistical dispersion of the data. Calculations such as the standard deviation (or multiples thereof), interquartile range, range or peak-to-peak, or noise calculation defined by ASTM can be used. Max Peak Height and Total Peak Area calculations are made for each region of interest, for example 521, defined by the user. For each region the Max Peak Height can be estimated by taking the maximum data point within the range. The Total Peak Area can be found by numerical integration of each region, resulting in an estimation of the total area of all peaks within those regions. Note that the random noise in the residual signal has little effect on numerical integration. This is because the noise has zero mean (removed during the decoupling operation) so the integration averages out the noise. These attributes, as well as other examples, are listed and defined in the table in FIG. 6b. The region of interest in FIG. 5c is designated 521.

[0043] Returning again to FIG. 4, the extracted attributes are delivered from attribute extractor 404 to decider 405. Also delivered to decider 405 are blank run acceptance criteria against which the extracted attributes are compared. The blank run acceptance criteria, determined at block 406, define the range in which the extracted attributes must fall, or the threshold or criteria that need to be met, for the chromatographic system to be considered viable. The user may load pre-defmed default blank run acceptance criteria or may manually set them, through UI 307 (FIG. 3) for example. In one embodiment, as illustrated by default blank run acceptance criteria storage block 407 in FIG 4, the user can prompt the system to load pre- defmed default blank run acceptance criteria based upon the characteristics of the GC method. For example, different blank run acceptance criteria could be defined for each unique detector type. The blank run acceptance criteria could be selected such that the evaluation works well with most GC methods using that detector type, GC column type, and/or the instrument conditions in the GC method. The default blank run acceptance criteria could be adjusted by the user in specialized cases where their GC method does not conform to the norm. In this case the default blank run acceptance criteria serve as guidance or as a good starting point for the user to adjust for their own needs.

[0044] The blank run acceptance criteria are thus called from block 406 and are applied to the extracted attributes from attribute extractor 404, which determines if the extracted attributes meet or violate the acceptance criteria, and, according to certain embodiments, if the system passes or fails the blank run analysis. In certain embodiments, the result of this comparison is output, providing an indication of whether the chromatographic system is viable (that is, performing normally and suitable for use), or if repair, warning or other measures are needed.

The interpretation of the result, indicating the need for repair, warning or other measures, can be provided by a reporter 408 and output to the user through UI 307, an automated email, text, or voicemail, or other means of communication.

[0045] In certain embodiments, if decider 405 determines that the system failed the blank run analysis (i.e. that the chromaotographic system is not viable), the system may choose to continue the sequence, pause the sequence to wait for user intervention, abort the sequence, or re-run the blank run. This option may be specified based on user input. The system can also suggest to the user a possible reason or reasons for failure and/or what may solve the issue that caused the failing condition. In some embodiments, the system could automatically fix the cause for failing the acceptance criteria. For example, the system could raise the oven temperature to clean out the instrument. The system could run an additional blank run after the potential fix is implemented to ascertain that the extracted attributes now meet the acceptance criteria.

[0046] In certain embodiments, if acceptance criteria are not provided for an attribute, the BRAU 400 assumes the user does not wish to use it for analyzing the blank run. The attribute extractor 404 may still calculate some or all attributes for its own history and, optionally, subsequent analysis by the user. However, it will not consider that attribute in evaluating the performance of the chromatographic system using the blank run. Additionally, if a user does not provide region(s) of interest for an attribute, the attribute extractor 404 may calculate that attribute using the entire chromatographic run time as the region of interest. Conversely, the system may simply store the blank run data and allow the user to reprocess it for blank run analysis at a later time and with the same or new acceptance criteria.

[0047] The tables in FIGS. 6a, 6b, and 6c list, by way of example, various attributes relating to blank runs in accordance with certain embodiments. Also listed are a description of the attributes, the failing conditions (acceptance criteria not being met) by which these attributes are evaluated where the max limit and min limit are defined by the acceptance criteria for example, and an interpretation of what may have caused the blank run to fail the acceptance criteria.

Certain attributes may only apply to solvent blank runs, such as those listed under‘Solvent Blank Run attributes in FIG. 6c.’

[0048] As seen from the table in FIG. 6a, considering the Initial Baseline attribute as an example, the baseline signal at initial conditions can indicate system contamination if the signal is above the acceptance criterion; or it can indicate a detector malfunction or a broken column if the signal is below the acceptance criterion. In either case, the instrument is not performing well enough to provide acceptable analyses. As another example, considering the Total Peak Area attribute in FIG. 6b, the summed peak area of the residual signal, can indicate inlet or gas contamination, carryover from a previous run, or column head fouling if the value is above the acceptance criterion.

[0049] In certain embodiments, storage and aggregation of analysis data, including, for example, extracted attributes from attribute extractor 404 and decisions from decider 405, may be desired. Extracted attributes and decision history storage 409 provides this functionality, storing the extracted attributes and decisions in memory 306 (FIG. 3) for example. Extracted attributes and decisions belonging to the same method, or method and solvent type, or any other grouping, can be grouped together and organized in files 309 (FIG. 3). More generally, the organization can be in the form of files, folders, directories, or other groupings, or any type of data structure or record. In certain embodiments, BRAU 400 can provide the user historical information regarding blank runs to help determine the health of the instrument, or can store the information so the user can analyze and infer its health. One application for this can include providing control charting for a number of runs before maintenance is needed.

[0050] FIG. 7 is a flow diagram showing an example use case in accordance with certain embodiments. At 701, the user creates a method with selected acceptance criteria. This can be input through UI 307 (FIG. 3). The user then runs the created method, at 702, either as a single analysis or as part of a sequence, with the run identified as a blank run, and, in certain

embodiments, as either a no-injection or solvent blank run. During or following the run, signal data is collected (at 703, from detector 303 in FIG. 3) and the blank run is evaluated (at 704) based on the acceptance criteria. A determination of whether the blank run is a“Pass” or a“Fail” is then made. If the determination is a Pass, the analysis is completed and the next run is awaited, or the instrument advances to the next line in the sequence of runs, for example. Other actions are also contemplated following a Pass. If, on the other hand, the determination is a Fail, the user may be notified, and a predetermined action may be performed. For instance, the instrument can “Pause” and await user intervention. Other options are for the instrument to rerun the blank, stop the sequence of runs, abort the sequence, perform a column bake out, or continue the sequence to the next sample.

[0051] While embodiments and applications have been shown and described, it would be apparent to those skilled in the art having the benefit of this disclosure that many more modifications than mentioned above are possible without departing from the inventive concepts disclosed herein. The invention, therefore, is not to be restricted except in the spirit of the appended claims.