Blank responses, variability

Assume that several aliquots of the unknown solution with a concentration r, are transferred to volumetric llasks having a volume V. To each of these flasks is added a variable volume V,t)f a standard solution of the analyte having a known concentration Suitable reagents are then added, and each solution is diluted to volume. Instrumental measurements are then made on eacii of these solutions and corrected for any blank response to yield a nei instrument response S. If the blank-corrected instrument rcspon.se is proportional to concentration, as is assumed in the standard-addition method, wc ma) write... 

A review of theory Indicates that the variability of blank responses Is the preferred basis for defining the lower limits of measurement. 

The variability of blank responses has been used In England to estimate the limits of detection for official methods of water analysis. In the United States, the variability of sample or standard responses has been more often used to estimate limits of detection. Practices also differ with respect to whether or not blank correction is done. These practices are compared and recommendations made regarding the most appropriate procedures for estimating lower limits of measurement for several types of environmental analyses. 

During the last 20 or so years there have been significant advances In the theory for defining the lower limits of measurement. Unfortunately, In practice we have not always applied what theory has told us. Theory tells us that the variability of blank responses Is the preferred basis for determining when the sample response Indicates that the determlnand concentration Is greater than zero. Heinrich Kaiser was one of the first scientists to recognize that fact, and he likened the situation to that of searching for a ship In a stormy sea. 

Several authors have published papers regarding the calculation of the limit of detection based on the variability of the blank responses. Prominent among these have been H. Kalser( ) In Germany, A.L. Wllson( ) In England, and L. Currle( 5) In the U.S.A. The following theoretical treatment Is based on the work of A.L. 

In summary. In defining the detection limit based on the variability of blank responses, one must make two choices. First, one must choose whether blank correction is to be based on a "well-known blank or on "paired-comparisons". Second, one must choose the values for a and B, corresponding to the risks of errors of the first and second kinds. Table I summarizes some of the definitions that have been proposed by Currie.(5)... 

Variability Of Blank Responses. In order to limit the discussion, let us focus on water analyses as representative of environmental analyses. In the United Kingdom, the Standing Committee of Analysts of the Department of the Environment issues analytical methods In a series of booklets. Included among these are the Methods for the Examination of Waters and Associated Materials. 10-14) Several of these methods have been evaluated by Individual laboratories to determine the limit of detection based on the variability of the blank and using paired comparisons for blank correction. Published values for the limit of detection for several of these methods are... 

The 99% confidence level corresponds mathematically to the case in which the error of the 1st kind Is chosen to be 1% (and the error of the 2nd kind Is equal to 50 %). However, the MDL Is based on the variability of analyte response rather than blank response. The MDL was first defined for application to the analysis of trace organic compounds and apparently was based on the conclusion that the first assumption listed above was not met. 

The preferred procedure for determining detection limit Is that based on the variability of blank responses for a complete analytical procedure. This... 

Standards and blanks are the usual controls used in analytical HPLC. Standards are usually interspersed with samples to demonstrate system performance over the course of a batch run. The successful run of standards before beginning analysis demonstrates that the system is suitable to use. In this way, no samples are run until the system is working well. Typically, standards are used to calculate column plate heights, capacity factors, and relative response factors. If day-to-day variability has been established by validation, the chromatographic system can be demonstrated to be within established control limits. One characteristic of good science is that samples... 

Optimization of System Variables. The dependence of the blank level and the total signal (blank + analyte response) on the liquid flow rate is shown in Figure 10. The conditions are the same as those for Figure 9 except that 10"6 M Hg2(N03)2 at pH 4 is used. Down to the lowest flow rate studied (1500 / L/min), the net response to 5 ppbv S02 is essentially constant. Unfortunately, this flow rate dependence was examined fairly late in the study and the other data reported here were obtained with a liquid flow rate of 2600 pL/min. It is clear, however, that down to at least 1500 nL/min, the response/blank ratio improves it may be advantageous to use a lower flow rate. This behavior also strongly suggests that the transport of mercury from the bulk solution (liberated due to the intrinsic disproprotionation equilibrium) to the carrier air stream is controlled by liquid phase mass transfer. 

Sensitivity. When operated in an ion monitoring mode as a GC detector, a mass spectrometer with standard electron multiplier detection and signal amplification is theoretically capable of producing a response to samples of less than 10 " mol (1 femtomole) [68]. In practice such limits have not been reached due to the combined effects of variable stationary phase and instrument background, sample degradation in gas chromatography or, in isotope dilution, residual blank contributions from... 

Responses are obtained from replicate analyses (n = 4) of a field blank and the field blank spiked to contain a level of analyte that is close to the LOQ. The means of the obtained responses are tested statistically (with the f-test) to determine whether they are statistically different. If the difference is significant, the variability of the response in the spiked sample is evaluated by taking the ratio of its mean response to the response standard deviation. If the ratio is greater than or equal to 3, then the spiked concentration is taken as the LOQ. 

Even with blank corrections, several factors can cause the basic assumption of the external standard method to break down. Matrix effects due to extraneous species in the sample that are not present in the standards or blank can cause the same analyte concentrations in the sample and standards to give different responses. Differences in experimental variables at the times at which blank, sample, and standard are measured can also invalidate the established calibration function. Even when the basic assumption is valid, errors can still occur owing to contamination during the sampling or sample preparation steps. 

The fuming HNO3 digestion procedure was used in conjunction with differential pulse anodic stripping voltammetry to determine sensitivity, linearity of response, measurement precision, and lead background or blank levels. Recovery studies, as well as some interference studies, were also conducted. It should be emphasized that the data presented in this discussion were obtained in a particular laboratory situation. Practical detection limits were determined both by the laboratory environment and by the ability of the analyst to reduce both the variability and magni-... 

For flame atomic absorption spectrophotometry, the detection limit Is defined as the concentration that produces absorption equivalent to twice the magnitude of the background fluctuation. No mention is made of the blank or blank correction. This definition implies an instrument detection limit rather than a detection limit of a complete analytical procedure. Finally, no mention Is made of the need to determine the variability of responses. 

Finally, there Is the attitude that equates the Instrument response to the analytical result. In this context, blank correction and the calculation of the variability of the blank Is not understood to be a necessary additional step. Too aoch emphasis has been placed on following certain rules, rather than doing what Is necessary to obtain the best possible estimate of the true value. 

The usual approach to answering either of these questions Is based on establishing the Instrument response with respect to analyte concentration and measuring the variability of this response when no analyte Is present. Some analysts prefer to call the measured response the blank. The solution used for the measurement Is also called the blank, and It Is the composition of this blank that often leads to the problems mentioned above. For simplicity of discussion, we will consider samples and blanks to be solutions. 

Limit of detection (LOD) is the lowest concentration of an analyte that the bioanalytical procedure can reliably differentiate from background noise. There are several approaches for determining the LOD (ICH Harmonized Tripartite Guideline, 2005), but a common practice is to evaluate the variability of the analytical background response of blank samples. To estimate the LOD, run blank (e.g., assay buffer, zero calibrator) sample replicates (>6) across one or more runs and calculate the mean background value 2 SD or 3SD to define the LOD. Although commonly used to define the sensitivity of an assay, LOD should be used with caution because the value is defined in an inherently variable region of the curve and is based upon a user-defined calculation. 

In analytical chemistry we always try to arrange that Equation [8.19a] provides an adeqnate model for the relationship between the instrumental response (Y) and the concentration (or amount) of analyte (x) injected directly into the instrument (instrumental cahbration) or used to spike a blank matrix (method calibration, see Section 8.5). When analytical chemists speak of a linear calibration equation they refer to Equation [8.19a], a simple linear regression model that is linear in both the fitting parameters and also in the independent variable Equation [8.19b] might be referred to as a non-linear calibration equation by a chemist, although to a statistician it is an example of a simple linear regression model, i.e., it is hnear in aU of the fitting parameters. 

Matrix Effect and Recovery For LC-MS/ MS-based methods, the signal suppression or enhancement of the analyte due to the presence of the matrix interferences (matrix effects) in MS/MS detection should be evaluated by comparing the response (peak area) of the analyte and the IS from the extracted blank samples post-fortified with the analyte and the IS with the response of neat solutions with both the analyte and the IS at the same concentrations as above. Matrix effects should be evaluated in one pooled batch of animal matrix or in at least three different batches of human matrix, using three replicates at a minimum of three QC concentrations (e.g., low quality control [LQC], medium quality control [MQC], high quality control [HQC]) with IS at working concentration. The coefficient of variation (CV%) of the matrix effect variability should be <15% at each concentration level and between the three (LQC, MQC, and HQC) concentration levels. 

So far we have been dealing with various forms of the "response" to displacements of atoms. In Section 6 also certain electric fields have been studied, we benefited from the fact that displacing atoms in GaAs generates dipoles and therefore electric fields all the reasonings of Section 6 were, however, limited to polar crystals and e.g. determination of static dielectric constant e as in Section 6.2 would be impossible in Ge or other homopolar substances. From the point of view of studying dielectric properties, the main drawback of Section 6 was our dependence upon the various displacement patterns the electric fields could not be varied at will, as an independent variable. The present Section summarizes the most recent applications of the DF which tend to fill this blank and to open the way to "direct" treatment of dielectric properties of semiconductors, within the framework of the Density Functional. They are the treatment of constant macroscopic electric field imposed from outside (Section 8.1) and "direct" evaluation. of the individual elements of the inverse dielectric matrix s ("q + +"g ) (Section 8.2). 

The method of standard addition is widely used, particularly when the composition of the sample matrix is variable or unknown so that the response of a reagent/matrix blank would be unreliable. At least three and preferably more spiked samples should be prepared, but if the amount of sample is limited, as few as one sample must suffice. It is especially useful with such analytical techniques as flame and plasma emission spectrometry and potentiometry (Topics E4, E5 and C8). 

---