Qualitative Aspects of Environmental Testing for Organic Parameters – An Overlooked Facet of Data Quality
By Lee Wolf, Regulatory Affairs Manager, ALS Environmental – USA
One may look at classical wet chemistry and the pioneering chemists and wonder. What were they really doing? What were they looking for? Discovering tests for isolating and identifying chemical components was the purpose. One cannot envision that accuracy and precision, or detection limits were of much concern. Fast-forward to the infancy of environmental testing and one can still find the identification of chemical components of utmost importance in developing methods and technology.
Currently, and for recent years, the focus of environmental measurements has predominately been on quantitative aspects. Significant emphasis is now put on detection limits, quantitation limits, accuracy and precision, and uncertainty of measurements. That is, an emphasis on sensitivity over selectivity, both in terms of data quality and data usability. Yet fundamentally, the targeted component(s) of a sample must first be accurately identified before accurate quantitation can occur.
In this paper, the shift in focus from selectivity to sensitivity in the analysis of organic compounds will be discussed and potentially overlooked qualitative data quality considerations examined. Examples of the reliance on standard methodologies and how they can lead to qualitative errors is examined, and how the need for a strong understanding of method selectivity and the use of qualitative tools is important.
Common Qualitative Procedures and Potential Errors
It is generally assumed that qualitative components of an analysis are considered and implemented as part of the laboratory’s methodology and operating procedures. Yet, Data Quality Objectives are often void of qualitative indicators. Although laboratories strive to implement the qualitative procedures included in a method, there are important, and in some cases critical, qualitative procedures that should be fully understood. This is also an important consideration as laboratories are developing more “in-house” procedures and employing modifications to published methods.
This condition can be influenced by more reliance on software and analytical technology advances, particularly by less experienced analysts. Analyst training requirements are often focused on the ability to follow a published procedure and yield acceptable quantitative results but are silent on demonstrated understanding of the chemistries and qualitative aspects of the methodology. In fact, some accreditation requirements for establishing analyst competency are based on meeting defined quantitative benchmarks.
How certain methods are applied can also be problematic. Methods for non-specific use are commonly developed using “clean” sample (or blank) matrices. They are in turn applied across a wider variety of matrices, some representing significant challenges.
Analyte extraction and separations – For extractable organics the extraction is generally not designed to isolate individual target parameters and thus the sample extraction provides little qualitative value. Co-extracted materials can result in interferences during instrumental analysis. An often-overlooked aspect of organics separations (extraction from the sample matrix) is the use of extract cleanups. Cleanups can be advantageous in further isolating target compounds – from both sample interferences and other co-extracted components. Techniques used for cleanups, for example silica gel and gel permeation chromatography, can significantly reduce interferences and improve selectivity at the instrument analysis step for individual compounds and for classes of compounds with similar chemical traits. For some target parameters and/or sample matrices, laboratories not employing cleanups may increase the potential for inaccurate qualitative determinations or extensively qualified data.
Chromatography – The science of chromatography is about separation and qualitative determination of parameters of interest and technology advances in the area of gas and liquid chromatography have been numerous over the last several years. Column phases have been developed to enhance separations and improve the ruggedness of the analysis. Yet, the fundamentals of chromatography cannot be ignored. The selection of GC or LC column phases must take into consideration the other chromatographic parameters and the characteristics of the target analytes to optimize separation. Using temperatures and flows that speed chromatography, fast GC, or the use of extremely small-bore columns to increase throughput may appear acceptable for analysis of known standards and clean samples, but may not work well for many samples. Reliable separation for various sample matrices and range of concentrations is the goal, which leads to improved qualitative data. Using marginal or deteriorated chromatography may result in marginal qualitative data, even with the use of mass spectrometer detectors.
Second-column confirmation is a key component to providing good qualitative data for methods not using mass spectrometry. The use of differing column phases between the two columns must yield substantially different compound elution patterns for dual column confirmation to be valuable, since detectors used are generally only selective to the class of compounds being analyzed. Identification errors can occur if the two columns to do provide significantly differing elution patterns or retention times. Optimized chromatography again plays a big part. If chromatographic performance is poor, co-eluting compounds or interferences can negatively impact qualitative data quality.
Mass spectrometry is often used in organics environmental analyses to help provide more dependable qualitative data and most methods are well-established. However, certain components of published methods should not be taken for granted, as they are essential to qualitative identification of sample constituents. Lack of proper tuning, mass calibration, or instrument configuration can potentially lead to errors.
Common environmental methods are well established and time-tested. But as previously discussed, some of the qualitative aspects may be easy to overlook. To generate rugged and accurate qualitative data analyst must understand the analysis from a systematic level. Understanding analysis parameters, instrument performance, and how they may be impacted by varying sample matrices, target parameters, and chemistries can be challenging. The following examples from two common analysis technologies help illustrate this point.
Organics by Gas Chromatography/Mass Spectrometry
The process of tuning the mass spectrometer is intended to establish known and consistent detector operation. This includes fragmentation, mass calibration, mass resolution, and achieving correct ion ratios. Requirements are built in to most methods to demonstrate acceptable tuning. Today, instrument manufacturers may incorporate “automatic” tuning capabilities into operating systems, yet the analyst must understand how the results of tuning may impact detection, ion ratios, and subsequent qualitative identification. Training to promote a more complete technical understanding of the mass spectrometer operation is a solid basis for evaluating sample data generated.
Understanding ionization and fragmentation is important in evaluating and using proper ions for identification of each compound of interest. Most environmental methods list the characteristic ions to be used for compound identification and prescribe criteria for relative intensities of the ions in order to identify the compounds. If the effects of scan rates, spectral interferences, or anomalies are not accounted for false positives or false negatives may result from strict application of these ion abundance criteria. For example, skewed spectra may result from use of improper scan rates and using skewed spectra for qualitative determinations could result in false positives or false negatives.
Chromatography is often an overlooked component of GC/MS analysis, with oftentimes too much reliance on the mass spectra to compensate for marginal chromatography. However, good chromatography with good peak resolution will help ensure good qualitative data. When choosing or accepting chromatographic parameters and run conditions the analyst must consider the impact on spectral quality and purity, which in turn affects qualitative data accuracy. Conversely, when chromatographic resolution and peak shape is weaker, spectral overlap and spectral interferences become more problematic. This becomes additionally difficult when complex sample matrices are being analyzed. A case in point illustrating the need for a combination of good chromatography and correct detector settings is the analysis of Polynuclear Aromatic Hydrocarbons (PAHs). Using common environmental analysis GC/MS instrumentation PAHs give very simple mass spectra with a predominant molecular ion response, a ion cluster around the molecular ion, and few (if any) other ions giving a useful relative response (although there are exceptions). PAHs generally elute as theory predicts, in order of boiling point, and this is tied to molecule size of the PAHS compound. It sounds simple – straightforward, predictable chromatography and spectra. However, the simplicity can be a hindrance as well. Making changes in chromatographic settings will have limited effect on improving elution characteristics, and the lack of characteristic masses with significant signal abundance limits identification options. In complex sample matrices containing PAHs, alkylated PAHs, hydrocarbons, and other compounds, the lack of characteristic ions hampers one’s ability to differentiate between PAHs with common mass fragments. In Selected Ion Monitoring (SIM) analysis, the lack of masses available to effectively monitor can potentially impact qualitative accuracy. It is essential that excellent chromatographic performance be maintained, through use of optimized parameters, instrument maintenance, and using effective sample cleanups. For SIM analysis, using as many characteristic masses as possible or practical will reduce the incorrect peak identification.
Organics by Gas Chromatography/Electron Capture Detector
The electron capture detector is used for routine pesticide and PCB analysis due to its selectivity towards compounds containing chlorine and fluorine atoms (as well as certain other electron-absorbing components). Its high sensitivity is useful in detecting very low amounts of target parameters. Although the ECD provides a significant measure of selectivity, chromatographic performance is essential to compound identification. Since a single GC/ECD analysis using one GC column does not yield definitive qualitative data the pesticide and PCB analyses typically employ a secondary column with differing peak elution characteristics, and provide a ‘confirmation’ of the primary column results. To obtain the best qualitative results, the laboratory must ensure that the confirmatory column results in sufficiently different elution characteristics and that chromatographic resolution on both columns is good.
As with most common gas chromatography methods that do not use mass spectrometer detectors, the GC/ECD methods can be subject to interferences despite detector selectivity. Laboratory quality control is used to control and minimize laboratory interferences, but sample interferences can pose a significant problem with qualitative identification. One example of particular note is the interference of PCBs on the determination of organochlorine pesticides. When performing the analysis for organochlorine pesticides using GC/ECD, the analyst must be aware of the potential for PCBs being present on the sample. As is commonly known, PCBs are made up of a broad range of individual PCB congeners, often in complex mixtures. On the GC/ECD this means many individual peaks. A myriad of PCB peaks eluting at or near the known retention times of targeted pesticides, on either or both GC columns, can lead to qualitative problems. In some instances, certain PCB congeners can be detected on both columns simultaneously within the retention time window of the target pesticide. For example, it has been demonstrated that the congeners making up PCB Aroclor 1254 do elute within retention time windows for 4,4’-DDT, 2,4’-DDT, and Endrin Aldehyde using columns commonly recommended for pesticide analysis. This method limitation can result in false positive detections. The laboratory must know and be aware of PCB interferences on pesticides.
Assuring Reliable and Accurate Qualitative Data
Users of environmental testing data are making decisions based on accurate and correct data coming from the laboratory. Increasingly, evaluation of the quality of the data is based only on adherence to methods and analysis protocols (validation) and on quantitative results (accuracy, precision, and uncertainty). The laboratory must focus not only on these components, but also ensure qualitative accuracy. Understanding how technical details and instrument operation details impacts detection and qualitative data is important. Here are some suggestions on how we can ensure qualitative data quality:
- Improved training. In many laboratories analyst training focuses on following the method being used, basic operation and maintenance of the instrument, and data evaluation. An analyst is considered as having qualified (or Demonstration of Capability) if they can perform replicate analysis meeting certain quantitative criteria. Training steps or Quality Assurance processes could be added to ensure the analyst has a technical understanding of qualitative factors and techniques.
- Improved methods. Most EPA chromatographic methods describe only the most basic steps to be taken to identify compounds under normal conditions and instrument operation. However, most of these methods given little guidance on how to make identifications using more complex data.
- More defined specifications for peak resolution, retention times, and GCMS-SIM. Some basic, standard requirements for acceptable chromatography and use of GCMS-SIM could help ‘raise the bar’ for the industry, without hampering productivity and innovation.
- Accreditation qualification provisions can be improved to incorporate a qualitative understanding component. Current accreditation standards, such as TNI, require only that the analyst be sufficiently trained. This is a very general requirement, and specific requirements for demonstration of capability use only quantitative evaluations.
- Qualitative DQOs. Can data validation be completely effective if there is little assessment of qualitative components? Although defining these DQOs would represent a challenge, having qualitative DQOs could be effective for some methods.
There is no disputing the importance of detection limits, quantitation limits, calibration validity, and quantitative accuracy in environmental measurements on which monitoring, remediation, and such decisions are made. But it is also important not to overlook the qualitative aspects, where a re-focus is warranted. This effort should not deter innovation by manufacturers of instruments and components, nor prevent laboratory production, but could lead to improved and more reliable data.