Design experimental

Experimental design is not directly related to chemical analysis, but it is important in that it determines the number of samples for processing. This could mean that there are too many tests for the laboratory to fit into its schedule, bearing in mind that there are many other customers clamouring for laboratory services. It could also mean that the cost is prohibitive for the funds available for the project. 

Some of the books on the design of scientific experiments appear far too theoretical for use in college field trials. However, three books in particular have proved useful in this Institute  

For example, the book by Gomez and Gomez describes many possible designs such as the Latin square and the lattice designs. The former can handle simultaneously two known sources of variation among experimental units. Chapters deal with Sampling in experimental plots, and the Presentation of research results.  

Experimental design tools are widely considered to be some of the most important tools in the development of chemometric models.20 At the same time, however, it is important to note that these tools can be effectively applied only in cases where sufficiently relevant 

There are a wide variety of experimental designs which can be used for different experimental objectives. Therefore, the first step in choosing an experimental design is to specify the experimental objective. In the application of chemometrics to PAC, the experimental objective most often consists of two parts  

Depending on the application, there could be additional, or different, objectives to the experiment. For example, one might not want to build a quantitative regression model, but rather identify which design variables affect the analyzer response the most. In this case, a set of tools called screening designs20 can be used to efficiently obtain the objective. 

Once the type of design is specified, sufficient information is available to generate a table containing a list of calibration samples, along with the values of each design variable for each sample. However, it is important to note that there are additional considerations which can further alter this experimental plan  

Design type Number of samples required (2 levels) Number of 

The experimental design is made using the following different input experimental factors (experimental conditions to maintain the positive and negative levels) amount of the compound studied (Xj) the degree of rubbing out of KBr pellets (X2)  

2235 cm peak (Strong Overlapped Frequency)—It is readily seen that all coefficients for the first signal are statistically nonsignificant. This means that no effect of influence of the input factors Xj could be found (the experimental error of repeatability is, in deed, higher that the calculated regression coefficients). 

1501 cm peak (Overlapping Frequency)—The behavior of this system is quite similar to the previous one. There are, however, some substantial differences. For instance, a4 is negative, which means that for this particular frequency the parameter X4 has an opposite effect as compared to frequency 2. The mixed effects are almost the same as in the previous system, which is, again, an indication of nonlinear interactions and synergetic effects of the single input parameters. 

The experimental design carried ont made it possible to determine in a qualitative way the impact of four input experimental factors (experimental conditions) on the spectroscopic signal for five frequencies. All the data obtained correlated with the validation discussed earlier. The following general conclusions can be offered  

It seems that this particular factor reveals no specific impact but rather a random one. 

In contrast to optimal design, Hamprecht and co-workers recently introdnced a space-filling design techniqne for compound selection. This stochastic method nses the best linear unbiased estimator, in the form of Kriging, to constrnct selection designs that optimize the integrated mean-square prediction error, or entropy. This 

Space-filling design attempts to sample chemical space more evenly than D-optimal design, when the response snrface is not uniform but irregular, as is often the case with heterogeneous compound collections. 

The second aspect of method validation is the experimental work. This typically involves initial experiments with analytical standards to confirm the reliability and repeatability of calibration of the system using only standards. The next step usually involves a series of analytical runs, conducted over several days or weeks, in which one or more analysts prepare calibration curves and analyze replicates of the typical analyte/matrix combinations and concentrations that are to be routinely analyzed using the method. The final phase of validation typically includes several runs in which fortified or incurred materials, again representing typical analyte/matrix combinations and concentrations, are provided blind to the analyst(s). The results are then summarized in a validation report, which again should receive appropriate peer review within the laboratory prior 

It is critical that this validation report be properly identified and available for review by external auditors. A properly designed validation exercise and the accompanying report can also form the basis for a publication in a scientific journal when the contents of the report are sufficiently novel. In subsequent sections of the chapter, we will discuss the different aspects of validation experiments typically conducted for methods intended for screening, quantification, or confirmation purposes. 

7 PERFORMANCE CHARACTERISTICS ASSESSED DURING METHOD DEVELOPMENT AND CONFIRMED DURING METHOD VALIDATION FOR QUANTITATIVE METHODS 

Although each CVD experiment is unique, some general comments related to designing CVD experiments may be presented. The ultimate design generally is an iterative procedure, and often borrows aspects from several processes. 

There are several key reasons why the chemist can be more productive if he or she understands the basis of design, including the following four main areas. 

Screening. These types of experiments involve seeing which factors are important for the success of a process. An example may be the study of a chemical reaction, dependent on proportion of solvent, catalyst concentration, temperature, pH, stirring rate, etc. Typically 10 or more factors might be relevant. Which can be eliminated, and which should be studied in detail Approaches such as factorial or Plackett-Burman designs (Sections 2.3.1-2.3.3) are useful in this context. 

Optimisation. This is one of the commonest applications in chemistry. How to improve a synthetic yield or a chromatographic separation Systematic methods can result in a better optimum, found more rapidly. Simplex is a classical method for optimisation (Section 2.6), although several designs such as mixture designs (Section 2.5) and central composite designs (Section 2.4) can also be employed to find optima. 

Quantitative modelling. Almost all experiments, ranging from simple linear calibration in analytical chemistry to complex physical processes, where a series of 

The reason for titis problem is that die infiuences of pH and temperature are not independent. In chemometric terms, they interact . In many cases, interactions are commonsense. The best pH in one solvent may be different to diat in another solvent. Chemistry is complex, but how to find the true optimum, by a quick and efficient manner, and be confident in die result Experimental design provides the chemist with a series of rules to guide the optimisation process which will be explored later. 

Calculating properties for a class of reagents need not he extremely fast because it must be performed only once for each class of starting materials. Thereafter, it can be used for the design of many libraries. 

Much experimental work has been carried out on ozonation in drinking, waste and process water treatment. And since there is still much to be learned about the mechanisms of ozonation, and many possibilities of utilizing its oxidizing potential many experiments will be carried out in the future. Not only researchers but also designers, manufacturers and users of ozonation systems will continue to do bench-scale testing because ozonation is so system dependent. Most full-scale applications have to be tried out bench-scale for each system considered. That means that there is a need for not only fundamental information about the mechanisms of ozonation, but also information on how to set-up experiments so that they produce results that can be interpreted and extrapolated. 

Good experimental design can help produce good results with a minimum of effort. There are of course always surprises and unexpected circumstances in the life of an experimenter, however, these can be minimized with good preparation, perhaps even turned to good use. 

Multispecies toxicity tests come in a wide variety of types (artificial streams, generic freshwater, simulated farm ponds, ditches, experimental plots, and forests), and they share basic properties. Experimental designs should take into account the advantage of these properties to ensure an interpretable experimental result. We propose the following design parameters for experimental design, analysis, and interpretation. 

Multispecies toxicity tests are complex structures. Complex structures are nonequilibrium, historical, and nonlinear. To measure the recovery of such a structure is to measure a property that does not exist for a complex structure. 

Multispecies toxicity tests are not repeatable in the strictest sense since each is sensitive to initial conditions. However, common patterns do appear, and these should be the focus of the investigation. 

All impacts can leave lasting effects. Therefore determination of a NOEC or LOEC is not warranted. 

In multispecies toxicity tests, the interactions among the component species should be understood. 

Impedance measurements are often used to identify physical phenomena that control an electrochemical reaction and to determine the corresponding physical properties. This chapter provides guidelines for the design of experimental cells, for selection of appropriate impedance parameters, and for selection of appropriate instrument controls. 

The classic texts on the design of experiments in scientific and engineering studies emphasize (1) measurement and instrumentation, (2) sources of error, (3) factor design etc. [30, 31] This section addresses step-by-step many of these issues for in situ studies, and does so by integrating relevant chemical and chemical engineering concerns and concepts. This section attempts to provide a very useful short-list of design considerations for the experimentalist so that Eq. (2) can be solved. 

The analytical method was validated at the LOQ (0.05mgkg ) for each analyte by satisfactory recoveries of the respective analytes from control samples that were fortified at the initiation of each analysis set. The fortified control samples were carried through the procedure with each analysis set. An analysis set consisted of a minimum of one control sample, one laboratory-fortified control sample, and several treated samples. 

A calibration curve was generated for each analyte at the initiation of the analytical phase of the smdy. Standard solutions for injection contained carfentrazone-ethyl or derivatized acid metabolites. Standard solutions were injected at the beginning of each set of assays and after every two or three samples to gage the instrument response. 

Mechanistic interpretation of inhibition data depends on the level of characterization. 

The default properties of CYP catalysis can be defined by Michaelis-Menten kinetics  

Competitive reversible inhibition results from inhibitor competing with substrate at the site of catalysis. In this instance, the rate is expressed by the following equation  

CYP2C9 CYP2C9 2 Arg144 — Cys144 Reduced activity [204] 

CYP2C19 CYP2C19i 2 Splicing defect Inactive enzyme [204] 

Methods of variance analysis are helpful tools to evaluate effects of factors on the results of experiments afterwards. On the other hand, it may be advantageous to plan experiments in a comparative way (comparative experiments). 

Statistical experimental design is characterized by the three basic principles Replication, Randomization and Blocking (block division, planned grouping). Latin square design is especially useful to separate nonrandom variations from random effects which interfere with the former. An example may be the identification of (slightly) different samples, e.g. sorts of wine, by various testers and at several days. To separate the day-to-day and/or tester-to-tester (laboratory-to-laboratory) variations from that of the wine sorts, an m x m Latin square design may be used. In case of m = 3 all three wine samples (a, b, c) are tested be three testers at three days, e.g. in the way represented in Table 5.8  

The results of the experiments are evaluated by means of three-way ANOVA in its simplest form, m = n = p and q = 1. The significance of the sample effect can principally be guaranteed also in the case that both testers and days have significant influence (Sharaf et al. [1986]). 

In contrast to common statistical techniques, by modern experimental design influencing factors are studied simultaneously (multifactorial design, MFD). The aim of MFD consists in an arrangement of factors in such a way that their influences can be quantified, compared and separated from random variations. 

Frequently the signal intensity of the analyte A is the target quantity, the influences on which are described by Eq. (3.16a). Handling all the influences (interferences and other factors) in the same way and holding xA at any constant value so that a0 = yAo + SAAxA, Eq. (3.16a) can be written 

Spectroelectrochemistry utilises the difference in the spectroscopic signature between the oxidised and the reduced form of a system to probe its redox properties. An incremental application of potential gradually changes the spectral profile of a system corresponding to the changes in the population of oxidised and reduced species. These spectral changes as a function of applied potential allow for the determination of various redox properties including, the 

Finding the optimum conditions for a chemical reaction is far from being a trivial task. In every case, there are numerous parameters which will affect the outcome. Moreover, these factors may pull in different directions. Different factors may also interact with each other in 

Simplex design, and the associated simplex procedure, is one of the many statistical methods for optimisation of reaction conditions. The concept is, as the name suggests, very straightforward and will be explained here in more detail since it is the easiest of the statistical techniques to explain and illustrate in a non-specialist text. 

To start the simplex, we set up a number of experiments. This number is equal to the number of variables plus 1. In our example, we have two variables and so we select three experiments to try. These are represented by A (7 min at 28 °C), (14 min at 36 °C) and C (10 min at 45 °C). We now run these experiments and measure the yield of each. The results of 65, 73 and 75%, respectively, tell us that we have probably not made a 

There are some things which the chemist must keep in mind when following the simplex procedure. For example, the procedure is very good at hill climbing but it will always climb the nearest hill. If there is more than one summit on the response surface and the first triangle is placed on the side of a hill which is not the highest, then it will climb the hill it is on and stop at that summit and will never see that a higher one exists. 

Statistical techniques can be employed to processes on production scale as well to those in the laboratory. However, on production scale, the changes to the reaction parameters will always be made in very small increments so as not to significantly jeopardise the yield and so each experiment will be repeated many times to ensure consistency before any conclusions are drawn. It is also essential to ensure that safety is never compromised as there are known cases where yield optimisation has transformed a safe process into a lethally dangerous one. 

Prior to enzyme kinetic investigation, the linearity of the relationship between the metabolite formation and the enzyme quantity and reaction time should be determined. The time courses and enzyme concentration dependencies of CYP activities are often linear when determined at concentration ranges for microsomal proteins of 0.2 2 mg/mL, although this should be confirmed for each metabolic pathway being studied. 

As the rates for CYP-mediated reactions are usually either low or moderate from a standard enzymology perspective, the reaction time for the linearity assay could be extended to 60 min, or longer. Six time points for such assays are common, for example, 0, 5, 15, 30, 45, and 60 min. 

After the testing for linearity (at typically 0.2 or 1 mg/mL of microsomal proteins and a 15- or 30-min period of incubation), experiments to determine other parameters can be designed. The rationale for a suitable assay condition is that such a condition will permit the formation of quantifiable amounts of the metabolites without markedly depleting (i.e., by less than 20%) the substrate. 

Estimates of and Ymax with reasonable accuracy probably require detection of the rates of metabolite formation with substrate concentrations spanning the range from 0.3 to 3 x is the apparent value of K.  

Unfortunately, the must be determined, since it is not readily predicted. The alternative, therefore, is the use of a default set of substrate concentrations, based largely on the general understanding of the kinetic characteristics of metabolic enzymes, particularly CYP-mediated reactions (Guengerich, 

Optimization covers many aspects in medicinal chemistry. The optimization of the affinity and often selectivity for the biological target and the pharmacokinetic properties of a lead compound are the primary goals of most preclinical research projects. Secondly, optimization strategies may be appUed to 

Various strategies have been advocated in order to cover the physicochemical parameter space of a series of new compounds as well as possible. Familiar strategies go back to proposals by Topliss and Craig. Both are schemes used for substituent variation at a selected site. The Topliss substitution scheme can be used to optimize aromatic and aliphatic substituents using a fixed set of substituents and rules. A Craig plot is a 2D plot of selected descriptors, for example, Hammett a (electronic properties) and Hansch 7T values (lipophilicity). Substituents can be selected from each quadrant of this plot such that they vary widely in their properties, for example, lipophilic and hydrophilic, electron-donor and electron-acceptor, and to ensure the two properties are not correlated in the selected set which is preferable for the generation of stable QSAR models. 

A further extension would be to consider a 3D Craig plot using three descriptors, for example, reflecting steric, lipophilic and electronic properties of the substituents. In that case, substituents may be chosen from the eight octants. If one wants to consider even more descriptors, this approach becomes impractical. In that case, more advanced experimental design techniques may be applied. One approach taken by Hansch and Leo was to use CA to define sets of aliphatic and aromatic substituents useful in the design of compounds for synthesis, such that various aspects of the substituents are taken into account in a balanced way.  

A large number of substituent descriptors have been reported in the literature. In order to use this information for substiment selection, appropriate statistical methods may be used. Pattern recognition or data reduction techniques, such as PCA or CA are good choices. As explained in Section in.B.3. in more detail, PCA consists of condensing the information in a data table into a few new descriptors made of linear combinations of the original ones. These new descriptors are called PCs or latent variables. This technique has been applied to define new descriptors for amino acids, as well as for aromatic or aliphatic substituents, which are called principal properties (PPs). These PPs can be used in FD methods or as variables in QSAR analysis.  

Zeolite Synthesis The synthesis of Na-Y zeolites is carried out in five steps. Fly ash is initially screened through a 355 pm mesh and the particles retained on the screen are put 

Catalysts Used for Removal of Various Substrates in CWPO 

Phenol LaTLj CUjOj perovskite catalyst (Sotelo et at, 2004) 

AlFe-plUared montmorillonite (Kiss et al., 2003) Fe-SBA-15 (Martinez et at, 200f7) 

CeOj-doped Fe203/y-Al203 (Liu et al., 2006) Al-Cu-pUlared clay (Kim et al., 2004) 

editor, Recommended Reference Materials for the Realization of Physicochemical Properties , Blackwell Scientific Publications. Oxford, 1987. 

Handbook of Organic Solvents , CRC Press, Boca Raton, FL, 1995. 

Riddick, W. B. Bunger, and T. K. Sakano, Organic Solvents , 4th Edition, Wiley, New York, 1986. 

Bailey, K. L. Chumey, and R. L. Nuttall, The NBS Tables of Chemical Thermodynamic Properties , J. Phys. Chem. Ref. Data. 1982, 11, Suppl. 2. 

---

It is essential for the rotating-disc that the flow remain laminar and, hence, the upper rotational speed of the disc will depend on the Reynolds number and experimental design, which typically is 1000 s or 10,000 rpm. On the lower lunit, 10 s or 100 rpm must be applied in order for the thickness of tlie boundary layer to be comparable to that of the radius of the disc. 

E J, J M Blaney, M A Siani, D C Spellmeyer, A K Wong and W H Moos 1995. Measuring fersity Experimental Design of Combinatorial Libraries for Drug Discovery. Journal of dicinal Chemistry 38 1431-1436. 

Each observation in any branch of scientific investigation is inaccurate to some degree. Often the accurate value for the concentration of some particular constituent in the analyte cannot be determined. However, it is reasonable to assume the accurate value exists, and it is important to estimate the limits between which this value lies. It must be understood that the statistical approach is concerned with the appraisal of experimental design and data. Statistical techniques can neither detect nor evaluate constant errors (bias) the detection and elimination of inaccuracy are analytical problems. Nevertheless, statistical techniques can assist considerably in determining whether or not inaccuracies exist and in indicating when procedural modifications have reduced them. 

Finding the End Point Potentiometrically Another method for locating the end point of a redox titration is to use an appropriate electrode to monitor the change in electrochemical potential as titrant is added to a solution of analyte. The end point can then be found from a visual inspection of the titration curve. The simplest experimental design (Figure 9.38) consists of a Pt indicator electrode whose potential is governed by the analyte s or titrant s redox half-reaction, and a reference electrode that has a fixed potential. A further discussion of potentiometry is found in Chapter 11. 

Although there are only three principal sources for the analytical signal—potential, current, and charge—a wide variety of experimental designs are possible too many, in fact, to cover adequately in an introductory textbook. The simplest division is between bulk methods, which measure properties of the whole solution, and interfacial methods, in which the signal is a function of phenomena occurring at the interface between an electrode and the solution in contact with the electrode. The measurement of a solution s conductivity, which is proportional to the total concentration of dissolved ions, is one example of a bulk electrochemical method. A determination of pH using a pH electrode is one example of an interfacial electrochemical method. Only interfacial electrochemical methods receive further consideration in this text. 

Electrochemical measurements are made in an electrochemical cell, consisting of two or more electrodes and associated electronics for controlling and measuring the current and potential. In this section the basic components of electrochemical instrumentation are introduced. Specific experimental designs are considered in greater detail in the sections that follow. 

Each of these experimental designs also uses a different type of instrument. To aid in understanding how they control and measure current and potential, these instruments are described as if they were operated manually. To do so the analyst... 

The experimental design for cathodic stripping voltammetry is similar to that for anodic stripping voltammetry with two exceptions. First, the deposition step in-... 

A one-factor-at-a-time optimization is consistent with a commonly held belief that to determine the influence of one factor it is necessary to hold constant all other factors. This is an effective, although not necessarily an efficient, experimental design when the factors are independent. Two factors are considered independent when changing the level of one factor does not influence the effect of changing the other factor s level. Table 14.1 provides an example of two independent factors. When factor B is held at level Bi, changing factor A from level Ai to level A2 increases the response from 40 to 80 thus, the change in response, AR, is... 

To develop an empirical model for a response surface, it is necessary to collect the right data using an appropriate experimental design. Two popular experimental designs are considered in the following sections. 

The experimental design for ruggedness testing is balanced in that each factor level is paired an equal number of times with the upper case and lower case levels... 

Experimental Design for a Ruggedness Test Involving Seven Factors... 

In the two-sample collaborative test, each analyst performs a single determination on two separate samples. The resulting data are reduced to a set of differences, D, and a set of totals, T, each characterized by a mean value and a standard deviation. Extracting values for random errors affecting precision and systematic differences between analysts is relatively straightforward for this experimental design. 

Oles, P. J. Fractional Factorial Experimental Design as a Teaching Tool for Quantitative Analysis, /. Chem. Educ. 

This experiment examines the effect of reaction time, temperature, and mole ratio of reactants on the synthetic yield of acetylferrocene by a Eriedel-Crafts acylation of ferrocene. A central composite experimental design is used to find the optimum conditions, but the experiment could be modified to use a factorial design. 

Note These equations are from Doming, S. N. Morgan, S. L. Experimental Design A Chemometric Approach. Elsevier Amsterdam, 1987, and pseudo-three-dimensional plots of the response surfaces can be found in their figures 11.4, 11.5, and 11.14. The response surface for problem (a) also is shown in Color Plate 13. 

A standard sample containing a known amount of analyte was carried through the procedure using the experimental design in Table 14.6. The percentage of the known amount of analyte actually found in the eight trials were found to be... 

Deming, S. N. Morgan, S. L. Experimental Design A Chemometric Approach. Elsevier Amsterdam, 1987. 

Hendrix, C. D. What Every Technologist Should Know About Experimental Design, Chemtech 1979, 9, 167-174. 

Morgan, E. Chemometrics Experimental Design. John Wiley and Sons Chichester, England, 1991. 

Armstrong, N. A. James, K. C. Pharmaceutical Experimental Design and Interpretation. Taylor and Francis London, 1996 as cited in Gonzalez, A. G. Anal. Chim. Acta 1998, 360, 227—241. 

---