Chapter 5 Quality Assurance and Calibration Methods

Overview Quality assurance Method validation External standards calibration Standard additions calibration Internal standards calibration

5-1: Quality Assurance Panel a shows results for the determination of lead in the same sample of river water. Of 181 labs, 18 reported results more than 50% above and four reported results more than 50% below the certified level of 62.3 ± 1.3 nM. Even though most labs in the study employed recognized quality management procedures, a large fraction of results did not include the certified range.

5-1: Quality Assurance Quality assurance is what we do to get the right answer for our purpose. Use objectives, from which specifications for data quality can be derived. Specifications could include requirements for sampling, accuracy, precision, specificity, detection limit, standards, and blank values. Assessment is the process of (1) collecting data to show that analytical procedures are operating within specified limits, and (2) verifying that final results meet use objectives.

5-1: Specifications Specifications could include: sampling requirements accuracy and precision rate of false results selectivity sensitivity acceptable blank values recovery of fortification (spike recovery) calibration checks quality control samples

5-1: Specifications Sampling requirements: Must collect representative samples, and analyte must be preserved after sample is collected. Otherwise, even the most accurate analysis is meaningless. Quality assurance begins with sampling. Accuracy and precision: Within practical restraints, (cost, time, and limited amounts of material), what level of accuracy and precision will satisfy the use objectives? What rate of false results is acceptable? A false positive implies that the concentration exceeds the legal limit when, in fact, the concentration is below the limit. A false negative implies that the concentration is below the limit when it is actually above the limit.

5-1: Specifications Selectivity (specificity): Sensitivity: distinguish analyte from other species in the sample (avoiding interference). Sensitivity: Capability of responding reliably and measurably to changes in analyte concentration The detection limit of an analytical method must be lower than the concentrations to be measured. Slope of the calibrations curve Example: method below is more sensitive for perchlorate in reagent water than in groundwater.

5-1: Specifications Acceptable blank values Blanks account for: Interference by other species in the sample Traces of analyte found in reagents used for sample preservation, preparation, and analysis Frequent measurements of blanks detect whether analyte from previous samples is carried into subsequent analyses by adhering to vessels or instruments. Three types – method, reagent, and field blanks.

5-1: Blanks Method blank: Reagent blank: Field blank: All components except analyte Taken through all steps of the analytical procedure Subtract the response of the method blank from the response of sample before calculating the quantity of analyte Reagent blank: Similar to a method blank, but it has not been subjected to all sample preparation procedures Field blank: Indicates if analyte is inadvertently picked up by exposure to field conditions

5-1: Matrix Interferences Sometimes, response to analyte can be decreased or increased by something else in the sample. Matrix refers to everything in the sample other than analyte.

5-1: Spike Recovery A spike, also called a fortification, is a known quantity of analyte added to a sample to test whether the response to the spike is the same as that expected from a calibration curve. Spiked samples are analyzed in the same manner as unknowns. If drinking water is found to contain 10.0 mg/L of nitrate, a spike of 5.0 mg/L could be added. Ideally, the concentration in the spiked portion found by analysis will be 15.0 mg/L. If a number other than 15.0 mg/L is found, then the matrix could be interfering with the analysis.

5-1: Spike Recovery

5-1: Specifications Calibration check: Quality control: Analyze solutions with known concentrations of analyte. A specification might, for example, call for one calibration check for every 10 samples. Calibration check solutions should be different from the ones used to prepare the original calibration curve. Helps verify that the initial calibration standards were made properly. Quality control: Samples of known composition are provided to the analyst as unknowns. Helps eliminate bias introduced by an analyst who knows the concentration of the calibration check sample.

5-1: Accuracy and Precision Accuracy can be assessed by: analyzing certified standards calibration checks performed by the analyst spikes made by the analyst analyzing blind quality control samples Gauge precision: replicate samples replicate portions of same sample

5-1: Standard Operating Procedures Written standard operating procedures (SOP) must be followed rigorously to avoid inadvertent changes in procedure that could affect the outcome. It is implicit that everyone follows the standard operating procedures. Adhering to these procedures guards against the normal human desire to take shortcuts based on assumptions that could be false.

5-1: Control Charts A control chart is a visual representation of confidence intervals for measurements having a Gaussian distribution. Warn us when a property being monitored strays dangerously far from an intended target value. Can be used to monitor accuracy, precision, or instrument performance as a function of time.

5-2: Method Validation Method validation is the process of proving that an analytical method is acceptable for its intended purpose. In validating a method, we typically demonstrate that requirements are met for specificity, linearity, accuracy, precision, range, limit of detection, limit of quantitation, and robustness.

5-2: Method Validation Specificity is the ability to distinguish analyte from anything else. Linearity is usually measured by the square of the correlation coefficient for the calibration curve. Types of precision include instrument precision, intra-assay precision, intermediate precision, and, most generally, interlaboratory precision. The “Horwitz trumpet” is an empirical statement that precision becomes poorer as analyte concentration decreases. Range is the concentration interval over which linearity, accuracy, and precision are acceptable.

5-2: Detection Limit (DL) The detection limit (LOD) is usually taken as three times the standard deviation of the blank. The lower limit of quantitation (LOQ) is 10 times the standard deviation of the blank. The reporting limit is the concentration below which regulations say that analyte is reported as “not detected,” even when it is observed. Robustness is the ability of an analytical method to be unaffected by small changes in operating parameters.

5-2: Detection Limit (DL)

5-3: Standard Addition A standard addition is a known quantity of analyte added to an unknown to increase the concentration of analyte. Standard additions are especially useful when matrix effects are important. Use Equation 5-7 to compute the quantity of analyte [X]i after a single standard addition. The initial concentration [X]i, increases to [X]f while the initial signal, Ix increase to Ix+s. [s]f is the concentration of standard added. Equation 5-7

5-3: Standard Addition

5-3: Multiple Standard Additions and Uncertainty For multiple standard additions to a single solution, use Equation 5-9 to construct the graph in Figure 5-6, in which the x-intercept gives us the concentration of analyte. For multiple solutions made up to the same final volume, the slightly different graph in Figure 5-7 is used. Equation 5-10 gives the x-intercept uncertainty in either graph.

5-3: Multiple Solutions Made up to the Same Final Volume For multiple solutions made up to the same final volume, the slightly different graph in Figure 5-7 is used.

5-3: Multiple Standard Addition to a Constant Volume

5-3: Uncertainty in the x-Intercept (ux) m = slope k = number of replicate measurements for unknown n = number of data points for calibration line 𝑦 = mean value of measured y for unknown x Sy = error of the regression

5-4: Internal Standard An internal standard is a known amount of a compound, different from analyte, that is added to the unknown. Signal from analyte is compared with signal from the internal standard to find out how much analyte is present. Internal standards are useful when the quantity of sample analyzed is not reproducible, when instrument response varies from run to run, or when sample losses occur in sample preparation. The response factor F is the relative response to analyte and standard.

5-4: Internal Standard Compensates for changes in experimental conditions, when Ix varies for same concentration [X] Internal Standard: Added standard is different from analyte

5-4: Internal Standards Calibration

5-4: Internal Standard For the sake of accuracy, it is best to prepare a series of standard mixtures of analyte plus internal standard and prepare a graph such as Figure 5-10. The response factor is the slope of the graph, and the intercept should be within statistical limits of zero. The graph should confirm linear response over the desired analytical range.

5-4: Internal Standard